Divergent effects of taxol on tumor necrosis factor-a-mediated cytolysis of ovarian carcinoma cells Sybilann Williams, MD, David G. Mutch, MD, Ling Xu, MD, and John Leslie Collins, PhD
St. Louis, Missouri OBJECTIVE: Our objective was to study the combined effect of taxol and tumor necrosis factor-a on the cytolysis of human ovarian carcinoma cell lines, because taxol has been shown to be active against ovarian carcinoma and has also been shown to increase tumor necrosis factor-a release from macrophages. STUDY DESIGN: The combined effect of taxol and tumor necrosis factor-a on the cell lines Caov-3, SK-OV-3, NIH: OVCAR-3, and A2780, which are sensitive to the cytolytic effect of tumor necrosis factor-a in the presence of inhibitors of protein synthesis, was investigated with a 24-hour chromium 51 release assay. RESULTS: At therapeutic concentrations taxol caused a significant increase in tumor necrosis factor-a-mediated cytolysis of Caov-3 and A2780 (p :5 0.05). By contrast, taxol caused a decrease in the tumor necrosis factor-a-mediated cytolysis of SK-OV-3 and NIH: OVCAR-3 (p :5 0.01). CONCLUSION: These results suggest that ovarian carcinomas have a heterogeneous response to the chemotherapeutic effect of taxol. (AM J OBSTET GVNECOL 1992; 167:1870-6.)
Key words: Taxol, tumor necrosis factor-a, cytolysis, ovarian carcinoma Ovarian carcinoma is the leading cause of death from gynecologic malignancy. Treatment for this disease is aimed at decreasing tumor burden, generally through surgery followed by chemotherapy. The introduction of cisplatin-based chemotherapy in the last decade has resulted in an increase in the frequency of response but little increase in the overall cure rate. 1 Significant problems remain with cisplatin-based therapy because of continued low survival, dose-limiting toxicities, and development of drug resistance. The need for new and more aggressive forms of therapy is clearly demonstrated by the fact that the age-adjusted death rate for women with ovarian carcinoma has remained constant for 35 years! Recently, taxol, a plant-derived antimitotic agent, has been undergoing intense study as a potential new drug in the treatment of epithelial ovarian cancer. Taxol is derived from extracts of the bark of the western yew tree (Taxus brevifolia); its complex taxane ring structure has so far eluded attempts at complete synthesis. Although other antimitotic chemotherapeutic agents disrupt microtubules, the unique antimitotic
From the Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine. Dr. Mutch was supported in part by an American Cancer Society Career Development Award. Presented in part at the Thirty-ninth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. Reprint requests: Sybilann Williams, MD, Department of Obstetrics and Gynecology, Washington University School of Medicine, 4911 Barnes Hospital Plaw, St. Louis, MO 63110. 6/6/41925
activity of taxol arises from its ability to stabilize microtubules, preventing their disassembly. 3 Phase II clinical trials of taxol in patients with drug-refractory epithelial ovarian cancer have shown response rates of 30% to 37%.3. 4 Phase III trials are presently ongoing. There are several lines of evidence that suggest that tumor necrosis factor-a (TNF -a) may be involved in host protective mechanisms, either as a macrophagesecreted product capable of mediating cytolysis directly5 or as a membrane-bound effector molecule associated with cytotoxic cells. 5 - 7 Additionally, there is evidence that patients with ovarian carcinomas (and carcinomas of other tissues) have significantly elevated serum levels of TNF -a, 8. 9 suggesting that this is a response to these cancers. Because taxol has been shown to be capable of increasing the release of TNF-a from macrophages, as well as decreasing the number of TNF-a receptors on some cancer cells,1O we investigated the effect of taxol on TNF-a-mediated cytolysis of human ovarian carcinoma cells in vitro. Material and methods Cell lines. The human ovarian carcinoma cells lines Caov-3, SK-OV-3, and NIH:OVCAR-3 (OVCAR-3) were obtained from American Type Culture Collection (ATCC, Rockville, Md.). The A2780 cell line was a kind gift of Drs. Alberto Manetta and Dennis Emma, University of California, Irvine Medical Center, Irvine, Calif. L929 (ATCC) is a murine fibroblast cell line known to be sensitive to TNF-a in vitro. All cells were maintained as exponential monolayer cultures in 100 mm tissue culture dishes (Corning Glass Works, Corning,
Effect of taxol on TNF-a-mediated cytolysis
Volume 167 Number 6
N.Y.) by passage twice a week. Cells for passage were removed with 0.05% trypsin (Gibco, Grand Island, N.Y.) in phosphate-buffered saline solution, pH 7.2 containing 0.04% ethylenediaminetetraacetate (Sigma Chemical Co., St. Louis) and 2 X 106 cells were plated per tissue culture dish. L929, Caov-3, and SK-OV-3 were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (HyClone Lab., Logan, Utah) and 30 mg/ml L-glutamine (Sigma). Cells were maintained in a humidifed atmosphere of 90% air and 10% carbon dioxide at 37° C. OVCAR-3 and A2780 cells were grown in RPMI-1640 medium supplemented with 10% fetal bovine serum, 30 mg/ml L-glutamine, and 10 f.1g/ml recombinant human insulin (Eli Lilly & Co., Indianapolis). Cells were maintained in a humidified atmosphere of 95% air and 5% carbon dioxide at 37° C. TNF-a. Recombinant human TNF-a was obtained as a frozen stock solution from Bachem California (Torrance, Calif.) at a concentration of 3.2 X 107 U/mg protein. The TNF-a was thawed and diluted in DMEM (L929, Caov-3, and SK-OV-3) or RPMI-1640 (OVCAR-3 and A2780) before its addition to cells. Taxol. Taxol (National Cancer Institute, Bethesda, Md.) was obtained as a stock solution of 6 mg/ml, diluted with DMEM to 2 X 10- 3 moVL, and stored at - 70° C. Taxol was thawed and diluted in appropriate medium immediately before its addition to cells. Protein synthesis inhibitors. Actinomycin D and emetine (both from Sigma) was stored at -70° C as stock solutions in medium used for routine growth of cells and further diluted in appropriate medium immediately before addition to cells. Assay of cytotoxicity. Cells were plated at a density of 2 x 106 cells per 100 mm plate in 10 ml of medium and allowed to adhere for 6 to 18 hours at 37° C. The medium was removed and replaced with 4 ml of RPMI1640 medium containing 0.3% bovine serum albumin (Sigma) and 50 f.1Cilml chromium 51 as sodium chromate (New England Nuclear/DuPont, Boston). Cells were labeled for 1 to 2 hours in a humidified atmosphere of 95% air and 5% carbon dioxide at 37° C, then washed with DMEM or RPMI-1640 medium as appropriate. Cells were removed from the plates with 0.04% ethylenediaminetetraacetate in phosphate-buffered saline solution, pH 7.2, and washed twice with appropriate medium. The cells were then resuspended, counted, and transferred to 96-well micro titer plates in a volume of 0.05 ml at a concentration of 2 X 105 cells/m!. Experimental wells received an additional 0.05 ml of taxol and control wells were given an additional 0.05 ml of medium. The plates were incubated for 2 hours at 37° C before addition of TNF-a and protein synthesis inhibitors or medium. Release of 5lCr was determined after 24 hours of incubation at 37° C in a humidified atmosphere of
1871
90% air and 10% carbon dioxide (L929, Caov-3, and SK-OV-3) or 95% air and 5% carbon dioxide (A2780 and OVCAR-3). Mter incubation the plates were centrifuged at 200g for 10 minutes. Aliquots of the supernatants were taken and counted in a 'Y-counter. Total incorporation of 5lCr was determined by counting the radioactivity of 105 cells. Three replicate wells were assayed for determination of specific lysis. The percent specific lysis ('ICr release) was determined by the following formula: Percent cytolysis = ~cpm 5lCr (experimental)] - [cpm 5lCr (spontaneous)]]
l-
[cpm 5lCr (total)] - [cpm 5lCr (spontaneous)] x 100
Assay of protein synthesis. Cells were plated at a concentration of 3 x 105 cells per well in 24-well plates and allowed to adhere overnight. The cells were then exposed to taxol or medium for 2 hours, then to emetine and/or TNF-a. Mter 2 hours, tritium-labeled leucine (New England Nuclear/DuPont) was added at a concentration of 10 f.1Cilml, and the cells were incubated for an additional 2 hours. The supernatants were removed and the reaction was terminated by the addition of 0.5 ml of 0.3 moVL sodium hydroxide. Mter vigorous mixing, 0.15 ml aliquots were removed and transferred to filter paper squares (Whatman No.5, Whatman International Ltd., Maidstone, England). Mter drying, the filters were washed three times in 15% trichloroacetic acid (Sigma) to precipitate protein. The filters were dried and counted in 5 ml of liquid scintillation fluid (Scintiverse II, Fisher Scientific, Fairlawn, N.J.). All wells were assayed in triplicate. Percent inhibition of protein synthesis was determined by the following formula: Percent inhibition = cpm tritium (experimental)] x 100 [1cpm tritium (control)
Statistical analysis. All wells were assayed in triplicate, and the experiments were repeated in quadruplicate to ensure reproducibility. The significance of treatment of cells with varying concentrations of taxol, TNF-a, and inhibitors of protein synthesis was determined with Student's t test. A significant difference was presumed to exist when p :5 0.05.
Results We have previously shown that all of the human ovarian carcinoma cell lines used in the current study express a resistance mechanism that blocks the cytolytic effect of TNF -a. Because the resistance mechanism is dependent on protein synthesis, these cells are relatively resistant to cytolysis by TNF-a in the absence of protein synthesis inhibitors but relatively sensitive to TNF -a-mediated cytolysis in the presence of protein
1872 Williams et al.
December 1992 Am J Obstet Gynecol
EMETINE
ACTINOMYCIN-D
NO PROTEIN SYNTHESIS INHIBITOR
100
100
100
...
.
....
. 80
80
80
~60
60
60
40
40
20
20
rn
in
~
0
I-
Z40 W
0
...«
a: w
Q.
.
« « «
20
o
0 ........_ .......'"
o
5
25
125
CONCENTRAllON OF TNF (UNITSlmL)
o
5
25
125
CONCENTRAllON OF TNF (UNITS/mL)
o
o
5
25
125
CONCENTRAllON OF TNF (UNlTS/mL)
Fig. 1. Percent cytolysis of Caov-3 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin-D (10- 8 moVL) or emetine (10- 6 moVL). There was a significant increase in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol (one asterisk, p :s 0.05; three asterisks, p :s 0.005).
synthesis inhibitors. 11 As shown in Fig. 1, taxol increases the TNF -a-mediated cytolysis of Caov-3 cells in the absence of protein synthesis inhibitors, as well as in the presence of emetine and to a lesser extent actinomycin D. Although taxol did not increase the TNF-a-mediated cytolysis of A2780 cells in the absence of protein synthesis inhibitors, it did increase the TNF-a-mediated cytolysis of these cells in the presence of emetine (Fig. 2). The cell lines OVCAR-3 and SK-OV-3 are similar to Caov-3 and A2780 in their tissue of origin (i.e., human ovarian carcinoma) and are also similar to Caov-3 and A2780 in that they are resistant to TNF-a-mediated cytolysis in the absence of protein synthesis inhibitors and sensitive to TNF-a-mediated cytolysis when protein synthesis is inhibited. II Despite these similarities, taxol did not increase TNF-a-mediated cytolysis of OVCAR-3 and SK-OV-3 cells; to the contrary, taxol caused a significant decrease in TNF -a-mediated cytolysis in the presence of emetine (Figs. 3 and 4). The decrease in TNF-a cytolysis seen in OVCAR-3 and SK-OV-3 when taxol was added 2 hours before the addition of TNF -a could potentially be explained by a
taxol-induced reduction in TNF -a receptor number or function, as previously described. 10 There was, however, no change in the level of inhibition ofTNF-a-mediated cytolysis when taxol was added 2 hours after the addition ofTNF-a (data not shown). Because OVCAR-3 and SK-OV-3 are insensitive to the cytolytic effect ofTNF-a in the absence of protein synthesis inhibitors, it is not possible to test the ability of taxol to decrease the cytolytic potential of TNF-a when protein synthesis inhibitors are not present. To determine whether the ability of taxol to decrease TNF -a-mediated cytolysis in OVCAR-3 and SK-OV-3 might be dependent on the concomitant inhibition of protein synthesis, we determined the effect of taxol on the TNF -a-mediated cytolysis of L929 cells. This transformed murine fibroblast cell line is sensitive to the cytolytic effect of TNF-a in the absence (as well as in the presence) of protein synthesis inhibitors. As shown in Fig. 5, taxol caused a decrease in TNF -a-mediated cytolysis in L929 irrespective of the absence or presence of inhibitors of protein synthesis. Because the level of TNF -a-mediated cytolysis is proportional to the level of inhibition of protein syn-
Effect of taxol on TNF-ot-mediated cytolysis
Volume 167 Number 6
NO PROTEIN SYNTHESIS INHIBITOR
ACTINOMYCIN-D
EMETINE
60
60
60
50
50
50
~4O
40
40
en
e ...!
o>30
..
30
.
1873
..
.
..
30
!Z
W
~20
20
20
10
10
10
W Q.
o~_....s.u.
o
5
25
125
CONCENlRAlION OF llIF (UNITSlmL)
o
o
5
25
125
CONCENTRATION OF llIF (UNITS/mL)
o
o
5
25
125
CONCENlRATlON OF llIF (UNITSJmL)
Fig. 2. Percent cytolysis of A2780 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitor actinomycin D (10- 8 moVL) or emetine (10- 6 moVL). There was a significant increase in percent TNF-ot-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of inhibitors of protein synthesis (one asterisk, p S 0.05; two asterisks, p S 0.01).
thesis, II it was possible that the increase of TNF-amediated cytolysis that results from the exposure of cells to taxol was related to the ability of taxol to inhibit protein synthesis. To test this hypothesis, protein synthesis was measured in Caov-3 and A2780 by the incorporation of tritium-labeled leucine in the presence of taxol, emetine, and TNF -a, either as single agents or in combination. As shown in Table I, taxol does inhibit protein synthesis in all the cell lines tested. Although the combination of taxol and TNF -a caused a significant increase in TNF -a-mediated cytolysis in Caov-3 cells when compared with cytolysis mediated by either TNF -a alone or taxol alone, this combination does not cause a significant increase in the inhibition of protein synthesis. Furthermore, although TNF -a-mediated cytolysis is increased in the presence of emetine, the additional increase in cytolysis that occurs in Caov-3 cells exposed to taxol, emetine, and TNF-a is not accompanied by an increase in the level of protein synthesis inhibition above that seen with emetine alone. An analogous situation exists for A2780 cells, in that the combination of taxol, emetine, and TNF-a increases cytolysis but not the level of protein synthesis inhibition,
when compared with emetine and TNF -a. These results indicate that the ability of taxol to increase the TNF-a-mediated cytolysis of Caov-3 or A2780 cells is not related to its ability to inhibit protein synthesis. The ability of taxol to inhibit protein synthesis also does not appear to be involved in its ability to inhibit TNF-a-mediated cytolysis in OVCAR-3, SK-OV-3, or L929 cells. As shown in Table I, taxol causes a significant decrease in the TNF -a-mediated cytolysis of OVCAR-3, SK-OV-3, and L929 cells in the presence of emetine without significantly altering the level of protein synthesis inhibition.
Comment Taxol increased the TNF-a-mediated cytolysis of Caov-3 and A2780 while decreasing the TNF-a-mediated cytolysis of OVCAR-3, SK-OV-3, and L929. There are several possible explanations for these results. The fact that taxol diminished TNF -a cytolysis in L929, OVCAR-3, and SK-OV-3 could potentially be explained by a mechanism whereby taxol decreased the number of TNF-a receptors on the cell surface, as has previously been described. 10 This would result in decreased inter-
1874 Williams et al.
December 1992 Am J Obstet Gynecol
EMETINE
ACTINOMYCIN-D
NO PROTEIN SYNTHESIS INHIBITOR 80
80
80
60
60
60
40
40
40
20
20
20
tn
en >
....
0
t ~
Z
LU (,)
a::
LU
a.
0 - ' - -......- ' 5 o 25
o
125
o
5
25
o
125
25
5
125
CONCENTRAllON OF TMF (UNITS/mL)
CONCENTRATION OF TMF (UNITSImL)
CONCENTRAllON OF TMF (UNITSImL)
o
Fig. 3. Percent cytolysis of OVCAR-3 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitor actinomycin D (10- 6 moVL) or emetine (10- 4 moVL). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of inhibitors of protein synthesis (one asterisk, p S 0.05; two asterisks, p S 0.01).
Table I. Effect of TNF-a, emetine, and taxol, as single agents or in combination, on protein synthesis in human ovarian carcinoma cells Cell line Caov-3
Treatment TNF-a* Emetind TaxoU: Emetine plus TNF-a Taxol plus TNF-a Taxol plus emetine Taxol plus TNF-a plus emetine
Cytolysis (%)
9 33 0 82 50 59 102
OVCAR-3
A2780
Protein synthesis inhibition (%)
-I
73 25 72 9 83 78
Cytolysis
Protein synthesis inhibition
2 16 0 43 3 29 58
-12 93 36 95 33 92 96
(%)
(%)
Cytolysis
Protein synthesis inhibition
2 36 0 74 4 38 50
5 98 48 98 37 99 99
(%)
(%)
SK-OV-3
L929
Cytolysis
Protein synthesis inhibition
Cytolysis
Protein synthesis inhibition
2 18 0 59 -2 -13 20
27 90 36 89 48 91 89
75 -7 0 62 21 5 16
3 93 36 93 29 81 83
(%)
(%)
(%)
(%)
*Concentration ofTNF-a was 125 Vim!' tConcentration of emetine was 10- 6 moVL for Caov-3 and A2780, 10- 5 moVL for L929, and 10- 4 moVL for OVCAR-3 and SK-OV-3. :):Concentration of taxol was 10- 5 moVL.
nalization of TNF-a and thereby diminished activation of the cytolytic mechanism. If taxol decreased TNF-a receptors in these cell lines and if the decrease was responsible for the decrease in TNF -a-mediated cytol-
ysis, then exposing these cells to TNF-a before their exposure to taxol should at least reduce the level of inhibition of TNF -a-mediated cytolysis by taxol. The fact that taxol was equally inhibitory whether it was
Effect of taxol on TNF-a-mediated cytolysis
Volume 167 Number 6
NO PROTEIN SYNTHESIS INHIBITOR
EMETINE
ACTINOMYCIN-D
60
60
60
40
40
40
tn
1875
en>
... to to
....I
0
!;:2O
...
0
20
20
0
0
Z W 0 D: W
a.
0
-20
...1..-_ _ _ _ _ __
o
5
25
125
CONCENTRAlION OF TNF (UNITS/mL)
-20 - ' - - - - - - - - o 5 25 125
CONCENTRAlION OFTNF (UNITSlmL)
-20 - ' - - - - - - - - o 5 25 125
CONCENTRAlION OF TNF (UNlTS/mL)
Fig_ 4. Percent cytolysis ofSK-OV-3 cells not pretreated (EZl) or pretreated (_) with 10- 6 moUL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin D (10- 5 moUL) or emetine (10- 4 moUL). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of emetine (three asterisks, p :s; 0.005; four asterisks, p :s; 0.001). added 2 hours before or 2 hours after TNF -a suggests that taxol does not decrease the TNF -a-mediated cytolysis of these cells by decreasing TNF -a receptors. Taxol caused an increase in TNF-a-mediated cytolysis in Caov-3 and A2780. The results seen in these cell lines cannot be due to a decline in either the number or the function of TNF-a receptors. Because TNF-a-mediated cytolysis in Caov-3 is increased by taxol in the absence of protein synthesis inhibitors, the possibility exists that taxol might be acting as a protein synthesis inhibitor. When protein synthesis was measured, it was demonstrated that at equivalent levels of TNF -a-mediated cytolysis in the presence of taxol or emetine, the levels of protein synthesis inhibition differed markedly (i.e., it was necessary to inhibit protein synthesis to a greater extent with emetine than with taxol to achieve similar levels of TNF -a-mediated cytolysis. Several studies 8 , 9, 11 have demonstrated increased circulating levels of TNF -a in the sera of patients with epithelial ovarian cancer. In addition, TNF-a messenger ribonucleic acid has been detected in situ in ovarian cancer. 12- 14 If TNF -a is involved in the surveillance of ovarian cancer, treatment with taxol could cause an alteration in the host response by acting synergistically or antagonistically with TNF-a, resulting in divergent effects on TNF-a-mediated cytolysis. In those patients
responsive to the combined effects treatment strategy may include administration of both taxol and TNF -a. It may prove to be of value to test the efficacy oftaxol in combination with TNF-a on cultured tumor cells from patients before instituting chemotherapy and to consider alternative therapy in those patients in whom diminished cytolysis is noted. REFERENCES 1. Omura GA, Bundy BN, Berek jS, Curry S, Delgado G, Mortel R. Randomized trial of cyclophosphamide plus cisplatin with or without doxorubicin in ovarian carcinoma: a gynecologic oncology group study. j Clin Oncol 1989;7:457-65. 2. DiSaia Pj, Creasman WT. Clinical gynecologic oncology. St Louis: CV Mosby, 1989:326. 3. McGuire WP, Rowinsky EK, Rosenshein NB, et al. Taxol: a unique antineoplastic agent with significant activity in advanced ovarian neoplasms. Ann Intern Med 1989; Ill: 273-9. 4. Thigpen T, Blessing j, Ball H, et al. Phase II trial of taxol as second-line therapy for ovarian carcinoma: a Gynecologic Oncology Group Study. Proc Am Soc Clin Oncol 1990;9:604. 5. Staren ED, Essner R, EconomoujS. Overview of biological response modifiers. Semin Surg Oncol 1989;5:379-84. 6. Warren jS, Ward PA, johnson KJ. Tumor necrosis factor: a pluripotential mediator of acute inflammation. Mod PathoI1988;1:242-7. 7. Wanebo HJ. Tumor necrosis factors. Semin Surg Oncol 1989;5:402-13.
1876 Williams et al.
December 1992 Am J Obstet Gynecol
NO PROTEIN SYNTHESIS INHIBITOR
EMETINE
ACTINOMYCIN-D
60
60
80
60
60
60
...J 40
40
40
en
en >
~ >
0
I-
ffi20
.... 20
20
0
0
.......
0
IX:
W
Q.
0
-20 - ' - - - - - - - - - o 5 25 125
CONCENTRATION OF TNF (UNITS/mL)
-20..L...-------25 125 o 5
CONCENTRATION OF TNF (UNITS/mL)
_20..L...-------5 25 125 o
CONCENTRATION OF TNF (UNlTS/mL)
Fig. 5. Percent cytolysis of L929 cells not pretreated (~) or pretreated (_) with 10- 6 mol!L taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin D (10- 5 mol!L) or emetine (10- 5 moI!L). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol (one asterisk, p ,.;; 0.05; three asterisks, p ,.;; 0.005; four asterisks, p ,.;; 0.001).
8. DeJaco P, Asselain B, Orlandi C, Fridman WH, Teillaud J-L. Evaluation of circulating tumor necrosis factor-a in patients with gynecological maliganacies. Int J Cancer 1991;4:375-8. 9. Balkwill F, Osborne R, Burke F, et al. Evidence for tumour necrosis factor/cachectin production in cancer. Lancet 1987;2:1229-32. 10. Ding AH, Porteu F, Sanchez E, Nathan CF. Shared actions of endotoxin and taxol on TNF receports and TNF release. Science 1990;248:370-2. 11. Powell CB, Mutch DG, Massad LS, Kao M-S, Collins JL. Common expression of a tumor necrosis factor resistance
mechanism among gynecological malignancies. Cancer Immunollmmunother 1990;32:131-6. 12. Spriggs DR, Imamura K, Rodriguez C, et al. Tumor necrosis factor expression in human epithelial tumor cells lines. J Clin Invest 1988;81:455-60. 13. Naylor MS, Malik STA, Stamp GWH, Jobling T, Balkwill FR. In situ detection of tumour necrosis factor in human ovarian cancer specimens. Eur J Cancer 1990;26: 1027 -30. 14. Takeyama H, Wakamiya N, O'Hara C, et al. Tumor necrosis factor expression by human ovarian carcinoma in vivo. Cancer Res 1991;51:4476-80.
St. Louis, Missouri OBJECTIVE: Our objective was to study the combined effect of taxol and tumor necrosis factor-a on the cytolysis of human ovarian carcinoma cell lines, because taxol has been shown to be active against ovarian carcinoma and has also been shown to increase tumor necrosis factor-a release from macrophages. STUDY DESIGN: The combined effect of taxol and tumor necrosis factor-a on the cell lines Caov-3, SK-OV-3, NIH: OVCAR-3, and A2780, which are sensitive to the cytolytic effect of tumor necrosis factor-a in the presence of inhibitors of protein synthesis, was investigated with a 24-hour chromium 51 release assay. RESULTS: At therapeutic concentrations taxol caused a significant increase in tumor necrosis factor-a-mediated cytolysis of Caov-3 and A2780 (p :5 0.05). By contrast, taxol caused a decrease in the tumor necrosis factor-a-mediated cytolysis of SK-OV-3 and NIH: OVCAR-3 (p :5 0.01). CONCLUSION: These results suggest that ovarian carcinomas have a heterogeneous response to the chemotherapeutic effect of taxol. (AM J OBSTET GVNECOL 1992; 167:1870-6.)
Key words: Taxol, tumor necrosis factor-a, cytolysis, ovarian carcinoma Ovarian carcinoma is the leading cause of death from gynecologic malignancy. Treatment for this disease is aimed at decreasing tumor burden, generally through surgery followed by chemotherapy. The introduction of cisplatin-based chemotherapy in the last decade has resulted in an increase in the frequency of response but little increase in the overall cure rate. 1 Significant problems remain with cisplatin-based therapy because of continued low survival, dose-limiting toxicities, and development of drug resistance. The need for new and more aggressive forms of therapy is clearly demonstrated by the fact that the age-adjusted death rate for women with ovarian carcinoma has remained constant for 35 years! Recently, taxol, a plant-derived antimitotic agent, has been undergoing intense study as a potential new drug in the treatment of epithelial ovarian cancer. Taxol is derived from extracts of the bark of the western yew tree (Taxus brevifolia); its complex taxane ring structure has so far eluded attempts at complete synthesis. Although other antimitotic chemotherapeutic agents disrupt microtubules, the unique antimitotic
From the Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Washington University School of Medicine. Dr. Mutch was supported in part by an American Cancer Society Career Development Award. Presented in part at the Thirty-ninth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. Reprint requests: Sybilann Williams, MD, Department of Obstetrics and Gynecology, Washington University School of Medicine, 4911 Barnes Hospital Plaw, St. Louis, MO 63110. 6/6/41925
activity of taxol arises from its ability to stabilize microtubules, preventing their disassembly. 3 Phase II clinical trials of taxol in patients with drug-refractory epithelial ovarian cancer have shown response rates of 30% to 37%.3. 4 Phase III trials are presently ongoing. There are several lines of evidence that suggest that tumor necrosis factor-a (TNF -a) may be involved in host protective mechanisms, either as a macrophagesecreted product capable of mediating cytolysis directly5 or as a membrane-bound effector molecule associated with cytotoxic cells. 5 - 7 Additionally, there is evidence that patients with ovarian carcinomas (and carcinomas of other tissues) have significantly elevated serum levels of TNF -a, 8. 9 suggesting that this is a response to these cancers. Because taxol has been shown to be capable of increasing the release of TNF-a from macrophages, as well as decreasing the number of TNF-a receptors on some cancer cells,1O we investigated the effect of taxol on TNF-a-mediated cytolysis of human ovarian carcinoma cells in vitro. Material and methods Cell lines. The human ovarian carcinoma cells lines Caov-3, SK-OV-3, and NIH:OVCAR-3 (OVCAR-3) were obtained from American Type Culture Collection (ATCC, Rockville, Md.). The A2780 cell line was a kind gift of Drs. Alberto Manetta and Dennis Emma, University of California, Irvine Medical Center, Irvine, Calif. L929 (ATCC) is a murine fibroblast cell line known to be sensitive to TNF-a in vitro. All cells were maintained as exponential monolayer cultures in 100 mm tissue culture dishes (Corning Glass Works, Corning,
Effect of taxol on TNF-a-mediated cytolysis
Volume 167 Number 6
N.Y.) by passage twice a week. Cells for passage were removed with 0.05% trypsin (Gibco, Grand Island, N.Y.) in phosphate-buffered saline solution, pH 7.2 containing 0.04% ethylenediaminetetraacetate (Sigma Chemical Co., St. Louis) and 2 X 106 cells were plated per tissue culture dish. L929, Caov-3, and SK-OV-3 were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (HyClone Lab., Logan, Utah) and 30 mg/ml L-glutamine (Sigma). Cells were maintained in a humidifed atmosphere of 90% air and 10% carbon dioxide at 37° C. OVCAR-3 and A2780 cells were grown in RPMI-1640 medium supplemented with 10% fetal bovine serum, 30 mg/ml L-glutamine, and 10 f.1g/ml recombinant human insulin (Eli Lilly & Co., Indianapolis). Cells were maintained in a humidified atmosphere of 95% air and 5% carbon dioxide at 37° C. TNF-a. Recombinant human TNF-a was obtained as a frozen stock solution from Bachem California (Torrance, Calif.) at a concentration of 3.2 X 107 U/mg protein. The TNF-a was thawed and diluted in DMEM (L929, Caov-3, and SK-OV-3) or RPMI-1640 (OVCAR-3 and A2780) before its addition to cells. Taxol. Taxol (National Cancer Institute, Bethesda, Md.) was obtained as a stock solution of 6 mg/ml, diluted with DMEM to 2 X 10- 3 moVL, and stored at - 70° C. Taxol was thawed and diluted in appropriate medium immediately before its addition to cells. Protein synthesis inhibitors. Actinomycin D and emetine (both from Sigma) was stored at -70° C as stock solutions in medium used for routine growth of cells and further diluted in appropriate medium immediately before addition to cells. Assay of cytotoxicity. Cells were plated at a density of 2 x 106 cells per 100 mm plate in 10 ml of medium and allowed to adhere for 6 to 18 hours at 37° C. The medium was removed and replaced with 4 ml of RPMI1640 medium containing 0.3% bovine serum albumin (Sigma) and 50 f.1Cilml chromium 51 as sodium chromate (New England Nuclear/DuPont, Boston). Cells were labeled for 1 to 2 hours in a humidified atmosphere of 95% air and 5% carbon dioxide at 37° C, then washed with DMEM or RPMI-1640 medium as appropriate. Cells were removed from the plates with 0.04% ethylenediaminetetraacetate in phosphate-buffered saline solution, pH 7.2, and washed twice with appropriate medium. The cells were then resuspended, counted, and transferred to 96-well micro titer plates in a volume of 0.05 ml at a concentration of 2 X 105 cells/m!. Experimental wells received an additional 0.05 ml of taxol and control wells were given an additional 0.05 ml of medium. The plates were incubated for 2 hours at 37° C before addition of TNF-a and protein synthesis inhibitors or medium. Release of 5lCr was determined after 24 hours of incubation at 37° C in a humidified atmosphere of
1871
90% air and 10% carbon dioxide (L929, Caov-3, and SK-OV-3) or 95% air and 5% carbon dioxide (A2780 and OVCAR-3). Mter incubation the plates were centrifuged at 200g for 10 minutes. Aliquots of the supernatants were taken and counted in a 'Y-counter. Total incorporation of 5lCr was determined by counting the radioactivity of 105 cells. Three replicate wells were assayed for determination of specific lysis. The percent specific lysis ('ICr release) was determined by the following formula: Percent cytolysis = ~cpm 5lCr (experimental)] - [cpm 5lCr (spontaneous)]]
l-
[cpm 5lCr (total)] - [cpm 5lCr (spontaneous)] x 100
Assay of protein synthesis. Cells were plated at a concentration of 3 x 105 cells per well in 24-well plates and allowed to adhere overnight. The cells were then exposed to taxol or medium for 2 hours, then to emetine and/or TNF-a. Mter 2 hours, tritium-labeled leucine (New England Nuclear/DuPont) was added at a concentration of 10 f.1Cilml, and the cells were incubated for an additional 2 hours. The supernatants were removed and the reaction was terminated by the addition of 0.5 ml of 0.3 moVL sodium hydroxide. Mter vigorous mixing, 0.15 ml aliquots were removed and transferred to filter paper squares (Whatman No.5, Whatman International Ltd., Maidstone, England). Mter drying, the filters were washed three times in 15% trichloroacetic acid (Sigma) to precipitate protein. The filters were dried and counted in 5 ml of liquid scintillation fluid (Scintiverse II, Fisher Scientific, Fairlawn, N.J.). All wells were assayed in triplicate. Percent inhibition of protein synthesis was determined by the following formula: Percent inhibition = cpm tritium (experimental)] x 100 [1cpm tritium (control)
Statistical analysis. All wells were assayed in triplicate, and the experiments were repeated in quadruplicate to ensure reproducibility. The significance of treatment of cells with varying concentrations of taxol, TNF-a, and inhibitors of protein synthesis was determined with Student's t test. A significant difference was presumed to exist when p :5 0.05.
Results We have previously shown that all of the human ovarian carcinoma cell lines used in the current study express a resistance mechanism that blocks the cytolytic effect of TNF -a. Because the resistance mechanism is dependent on protein synthesis, these cells are relatively resistant to cytolysis by TNF-a in the absence of protein synthesis inhibitors but relatively sensitive to TNF -a-mediated cytolysis in the presence of protein
1872 Williams et al.
December 1992 Am J Obstet Gynecol
EMETINE
ACTINOMYCIN-D
NO PROTEIN SYNTHESIS INHIBITOR
100
100
100
...
.
....
. 80
80
80
~60
60
60
40
40
20
20
rn
in
~
0
I-
Z40 W
0
...«
a: w
Q.
.
« « «
20
o
0 ........_ .......'"
o
5
25
125
CONCENTRAllON OF TNF (UNITSlmL)
o
5
25
125
CONCENTRAllON OF TNF (UNITS/mL)
o
o
5
25
125
CONCENTRAllON OF TNF (UNlTS/mL)
Fig. 1. Percent cytolysis of Caov-3 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin-D (10- 8 moVL) or emetine (10- 6 moVL). There was a significant increase in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol (one asterisk, p :s 0.05; three asterisks, p :s 0.005).
synthesis inhibitors. 11 As shown in Fig. 1, taxol increases the TNF -a-mediated cytolysis of Caov-3 cells in the absence of protein synthesis inhibitors, as well as in the presence of emetine and to a lesser extent actinomycin D. Although taxol did not increase the TNF-a-mediated cytolysis of A2780 cells in the absence of protein synthesis inhibitors, it did increase the TNF-a-mediated cytolysis of these cells in the presence of emetine (Fig. 2). The cell lines OVCAR-3 and SK-OV-3 are similar to Caov-3 and A2780 in their tissue of origin (i.e., human ovarian carcinoma) and are also similar to Caov-3 and A2780 in that they are resistant to TNF-a-mediated cytolysis in the absence of protein synthesis inhibitors and sensitive to TNF-a-mediated cytolysis when protein synthesis is inhibited. II Despite these similarities, taxol did not increase TNF-a-mediated cytolysis of OVCAR-3 and SK-OV-3 cells; to the contrary, taxol caused a significant decrease in TNF -a-mediated cytolysis in the presence of emetine (Figs. 3 and 4). The decrease in TNF-a cytolysis seen in OVCAR-3 and SK-OV-3 when taxol was added 2 hours before the addition of TNF -a could potentially be explained by a
taxol-induced reduction in TNF -a receptor number or function, as previously described. 10 There was, however, no change in the level of inhibition ofTNF-a-mediated cytolysis when taxol was added 2 hours after the addition ofTNF-a (data not shown). Because OVCAR-3 and SK-OV-3 are insensitive to the cytolytic effect ofTNF-a in the absence of protein synthesis inhibitors, it is not possible to test the ability of taxol to decrease the cytolytic potential of TNF-a when protein synthesis inhibitors are not present. To determine whether the ability of taxol to decrease TNF -a-mediated cytolysis in OVCAR-3 and SK-OV-3 might be dependent on the concomitant inhibition of protein synthesis, we determined the effect of taxol on the TNF -a-mediated cytolysis of L929 cells. This transformed murine fibroblast cell line is sensitive to the cytolytic effect of TNF-a in the absence (as well as in the presence) of protein synthesis inhibitors. As shown in Fig. 5, taxol caused a decrease in TNF -a-mediated cytolysis in L929 irrespective of the absence or presence of inhibitors of protein synthesis. Because the level of TNF -a-mediated cytolysis is proportional to the level of inhibition of protein syn-
Effect of taxol on TNF-ot-mediated cytolysis
Volume 167 Number 6
NO PROTEIN SYNTHESIS INHIBITOR
ACTINOMYCIN-D
EMETINE
60
60
60
50
50
50
~4O
40
40
en
e ...!
o>30
..
30
.
1873
..
.
..
30
!Z
W
~20
20
20
10
10
10
W Q.
o~_....s.u.
o
5
25
125
CONCENlRAlION OF llIF (UNITSlmL)
o
o
5
25
125
CONCENTRATION OF llIF (UNITS/mL)
o
o
5
25
125
CONCENlRATlON OF llIF (UNITSJmL)
Fig. 2. Percent cytolysis of A2780 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitor actinomycin D (10- 8 moVL) or emetine (10- 6 moVL). There was a significant increase in percent TNF-ot-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of inhibitors of protein synthesis (one asterisk, p S 0.05; two asterisks, p S 0.01).
thesis, II it was possible that the increase of TNF-amediated cytolysis that results from the exposure of cells to taxol was related to the ability of taxol to inhibit protein synthesis. To test this hypothesis, protein synthesis was measured in Caov-3 and A2780 by the incorporation of tritium-labeled leucine in the presence of taxol, emetine, and TNF -a, either as single agents or in combination. As shown in Table I, taxol does inhibit protein synthesis in all the cell lines tested. Although the combination of taxol and TNF -a caused a significant increase in TNF -a-mediated cytolysis in Caov-3 cells when compared with cytolysis mediated by either TNF -a alone or taxol alone, this combination does not cause a significant increase in the inhibition of protein synthesis. Furthermore, although TNF -a-mediated cytolysis is increased in the presence of emetine, the additional increase in cytolysis that occurs in Caov-3 cells exposed to taxol, emetine, and TNF-a is not accompanied by an increase in the level of protein synthesis inhibition above that seen with emetine alone. An analogous situation exists for A2780 cells, in that the combination of taxol, emetine, and TNF-a increases cytolysis but not the level of protein synthesis inhibition,
when compared with emetine and TNF -a. These results indicate that the ability of taxol to increase the TNF-a-mediated cytolysis of Caov-3 or A2780 cells is not related to its ability to inhibit protein synthesis. The ability of taxol to inhibit protein synthesis also does not appear to be involved in its ability to inhibit TNF-a-mediated cytolysis in OVCAR-3, SK-OV-3, or L929 cells. As shown in Table I, taxol causes a significant decrease in the TNF -a-mediated cytolysis of OVCAR-3, SK-OV-3, and L929 cells in the presence of emetine without significantly altering the level of protein synthesis inhibition.
Comment Taxol increased the TNF-a-mediated cytolysis of Caov-3 and A2780 while decreasing the TNF-a-mediated cytolysis of OVCAR-3, SK-OV-3, and L929. There are several possible explanations for these results. The fact that taxol diminished TNF -a cytolysis in L929, OVCAR-3, and SK-OV-3 could potentially be explained by a mechanism whereby taxol decreased the number of TNF-a receptors on the cell surface, as has previously been described. 10 This would result in decreased inter-
1874 Williams et al.
December 1992 Am J Obstet Gynecol
EMETINE
ACTINOMYCIN-D
NO PROTEIN SYNTHESIS INHIBITOR 80
80
80
60
60
60
40
40
40
20
20
20
tn
en >
....
0
t ~
Z
LU (,)
a::
LU
a.
0 - ' - -......- ' 5 o 25
o
125
o
5
25
o
125
25
5
125
CONCENTRAllON OF TMF (UNITS/mL)
CONCENTRATION OF TMF (UNITSImL)
CONCENTRAllON OF TMF (UNITSImL)
o
Fig. 3. Percent cytolysis of OVCAR-3 cells not pretreated (~) or pretreated (_) with 10- 6 moVL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitor actinomycin D (10- 6 moVL) or emetine (10- 4 moVL). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of inhibitors of protein synthesis (one asterisk, p S 0.05; two asterisks, p S 0.01).
Table I. Effect of TNF-a, emetine, and taxol, as single agents or in combination, on protein synthesis in human ovarian carcinoma cells Cell line Caov-3
Treatment TNF-a* Emetind TaxoU: Emetine plus TNF-a Taxol plus TNF-a Taxol plus emetine Taxol plus TNF-a plus emetine
Cytolysis (%)
9 33 0 82 50 59 102
OVCAR-3
A2780
Protein synthesis inhibition (%)
-I
73 25 72 9 83 78
Cytolysis
Protein synthesis inhibition
2 16 0 43 3 29 58
-12 93 36 95 33 92 96
(%)
(%)
Cytolysis
Protein synthesis inhibition
2 36 0 74 4 38 50
5 98 48 98 37 99 99
(%)
(%)
SK-OV-3
L929
Cytolysis
Protein synthesis inhibition
Cytolysis
Protein synthesis inhibition
2 18 0 59 -2 -13 20
27 90 36 89 48 91 89
75 -7 0 62 21 5 16
3 93 36 93 29 81 83
(%)
(%)
(%)
(%)
*Concentration ofTNF-a was 125 Vim!' tConcentration of emetine was 10- 6 moVL for Caov-3 and A2780, 10- 5 moVL for L929, and 10- 4 moVL for OVCAR-3 and SK-OV-3. :):Concentration of taxol was 10- 5 moVL.
nalization of TNF-a and thereby diminished activation of the cytolytic mechanism. If taxol decreased TNF-a receptors in these cell lines and if the decrease was responsible for the decrease in TNF -a-mediated cytol-
ysis, then exposing these cells to TNF-a before their exposure to taxol should at least reduce the level of inhibition of TNF -a-mediated cytolysis by taxol. The fact that taxol was equally inhibitory whether it was
Effect of taxol on TNF-a-mediated cytolysis
Volume 167 Number 6
NO PROTEIN SYNTHESIS INHIBITOR
EMETINE
ACTINOMYCIN-D
60
60
60
40
40
40
tn
1875
en>
... to to
....I
0
!;:2O
...
0
20
20
0
0
Z W 0 D: W
a.
0
-20
...1..-_ _ _ _ _ __
o
5
25
125
CONCENTRAlION OF TNF (UNITS/mL)
-20 - ' - - - - - - - - o 5 25 125
CONCENTRAlION OFTNF (UNITSlmL)
-20 - ' - - - - - - - - o 5 25 125
CONCENTRAlION OF TNF (UNlTS/mL)
Fig_ 4. Percent cytolysis ofSK-OV-3 cells not pretreated (EZl) or pretreated (_) with 10- 6 moUL taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin D (10- 5 moUL) or emetine (10- 4 moUL). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol, in presence of emetine (three asterisks, p :s; 0.005; four asterisks, p :s; 0.001). added 2 hours before or 2 hours after TNF -a suggests that taxol does not decrease the TNF -a-mediated cytolysis of these cells by decreasing TNF -a receptors. Taxol caused an increase in TNF-a-mediated cytolysis in Caov-3 and A2780. The results seen in these cell lines cannot be due to a decline in either the number or the function of TNF-a receptors. Because TNF-a-mediated cytolysis in Caov-3 is increased by taxol in the absence of protein synthesis inhibitors, the possibility exists that taxol might be acting as a protein synthesis inhibitor. When protein synthesis was measured, it was demonstrated that at equivalent levels of TNF -a-mediated cytolysis in the presence of taxol or emetine, the levels of protein synthesis inhibition differed markedly (i.e., it was necessary to inhibit protein synthesis to a greater extent with emetine than with taxol to achieve similar levels of TNF -a-mediated cytolysis. Several studies 8 , 9, 11 have demonstrated increased circulating levels of TNF -a in the sera of patients with epithelial ovarian cancer. In addition, TNF-a messenger ribonucleic acid has been detected in situ in ovarian cancer. 12- 14 If TNF -a is involved in the surveillance of ovarian cancer, treatment with taxol could cause an alteration in the host response by acting synergistically or antagonistically with TNF-a, resulting in divergent effects on TNF-a-mediated cytolysis. In those patients
responsive to the combined effects treatment strategy may include administration of both taxol and TNF -a. It may prove to be of value to test the efficacy oftaxol in combination with TNF-a on cultured tumor cells from patients before instituting chemotherapy and to consider alternative therapy in those patients in whom diminished cytolysis is noted. REFERENCES 1. Omura GA, Bundy BN, Berek jS, Curry S, Delgado G, Mortel R. Randomized trial of cyclophosphamide plus cisplatin with or without doxorubicin in ovarian carcinoma: a gynecologic oncology group study. j Clin Oncol 1989;7:457-65. 2. DiSaia Pj, Creasman WT. Clinical gynecologic oncology. St Louis: CV Mosby, 1989:326. 3. McGuire WP, Rowinsky EK, Rosenshein NB, et al. Taxol: a unique antineoplastic agent with significant activity in advanced ovarian neoplasms. Ann Intern Med 1989; Ill: 273-9. 4. Thigpen T, Blessing j, Ball H, et al. Phase II trial of taxol as second-line therapy for ovarian carcinoma: a Gynecologic Oncology Group Study. Proc Am Soc Clin Oncol 1990;9:604. 5. Staren ED, Essner R, EconomoujS. Overview of biological response modifiers. Semin Surg Oncol 1989;5:379-84. 6. Warren jS, Ward PA, johnson KJ. Tumor necrosis factor: a pluripotential mediator of acute inflammation. Mod PathoI1988;1:242-7. 7. Wanebo HJ. Tumor necrosis factors. Semin Surg Oncol 1989;5:402-13.
1876 Williams et al.
December 1992 Am J Obstet Gynecol
NO PROTEIN SYNTHESIS INHIBITOR
EMETINE
ACTINOMYCIN-D
60
60
80
60
60
60
...J 40
40
40
en
en >
~ >
0
I-
ffi20
.... 20
20
0
0
.......
0
IX:
W
Q.
0
-20 - ' - - - - - - - - - o 5 25 125
CONCENTRATION OF TNF (UNITS/mL)
-20..L...-------25 125 o 5
CONCENTRATION OF TNF (UNITS/mL)
_20..L...-------5 25 125 o
CONCENTRATION OF TNF (UNlTS/mL)
Fig. 5. Percent cytolysis of L929 cells not pretreated (~) or pretreated (_) with 10- 6 mol!L taxol in absence of protein synthesis inhibitors or in presence of protein synthesis inhibitors actinomycin D (10- 5 mol!L) or emetine (10- 5 moI!L). There was a significant decrease in percent TNF-a-mediated cytolysis of cells pretreated with taxol as compared with cells not pretreated with taxol (one asterisk, p ,.;; 0.05; three asterisks, p ,.;; 0.005; four asterisks, p ,.;; 0.001).
8. DeJaco P, Asselain B, Orlandi C, Fridman WH, Teillaud J-L. Evaluation of circulating tumor necrosis factor-a in patients with gynecological maliganacies. Int J Cancer 1991;4:375-8. 9. Balkwill F, Osborne R, Burke F, et al. Evidence for tumour necrosis factor/cachectin production in cancer. Lancet 1987;2:1229-32. 10. Ding AH, Porteu F, Sanchez E, Nathan CF. Shared actions of endotoxin and taxol on TNF receports and TNF release. Science 1990;248:370-2. 11. Powell CB, Mutch DG, Massad LS, Kao M-S, Collins JL. Common expression of a tumor necrosis factor resistance
mechanism among gynecological malignancies. Cancer Immunollmmunother 1990;32:131-6. 12. Spriggs DR, Imamura K, Rodriguez C, et al. Tumor necrosis factor expression in human epithelial tumor cells lines. J Clin Invest 1988;81:455-60. 13. Naylor MS, Malik STA, Stamp GWH, Jobling T, Balkwill FR. In situ detection of tumour necrosis factor in human ovarian cancer specimens. Eur J Cancer 1990;26: 1027 -30. 14. Takeyama H, Wakamiya N, O'Hara C, et al. Tumor necrosis factor expression by human ovarian carcinoma in vivo. Cancer Res 1991;51:4476-80.
