Anti-proliferative Effect on Human Pancreatic Cancer Cells of Natural Human Tumour Necrosis Factor-B Combined with Natural Human Interferon-a or Interferon-y T
WATANABE,
S
FUCHIMOTO,
N
MATSUBARA,
H
IWAGAKI AND
K
ORITA
First Department of Surgery, Okayama University Medical School, Okayama, Japan
The anti-proliferative effects of natural cytokines, human tumour necrosis factor-B. natural human interferon-a and natural human interferon-y, on three human pancreatic cancer cell lines (PANC-l, MIA PaCa-2 and BxPC-3) were investigated in vitro. The anti-proliferative effect was determined using the dye uptake method and analysed for synergism by the median effect principle. Tumour necrosis factor-B. as a single agent, had little anti-proliferative effect on any of the three cell lines, whereas interferon-a and interferon-y exhibited a strong anti-proliferative effect against two cell lines (MIA PaCa-2 and BxPC-3) and one cell line (BxPC-3), respectively. When tumour necrosis factor-B and interferon-a were administered together (ratio 1: 1), a synergistic effect was observed against PANC-l cells. The combination of tumour necrosis factor-f and interferon-y (ratio 10: 1) was synergisic against both PANC-l and MIA PaCa-2 cells. A synergistic anti-proliferative effect oftumour necrosis factor-B and interferons was, therefore, observed even for cell lines that showed little biological response to
112 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et at. Combination cytokine therapy
each cytokine alone. The data suggest that some future improvement in the treatment of pancreatic cancer may be obtained by using combination cytokine therapy. KEY WORDS:
TUMOUR NECROSIS
FACTOR-P;
INTERFERON-a;
INTERFERON-y; CYTOKINES; HUMAN PANCREATIC CANCER; IMMUNOTHERAPY
MATERIALS AND METHODS
INTRODUCTION Pancreatic cancer has one of the worst prognoses among all the malignancies and, despite many efforts to improve the therapy available for this tumour, no satisfactory progress has been achieved. Effective adjuvant therapy, such as radiotherapy, chemotherapy, immunotherapy, hyperthermia or hormonal therapy, is required to supplement surgery. The efficacy of several biological response modifiers for pancreatic cancer has been investigated, but most have produced little response. It has been reported that the treatment of various tumour cell lines with tumour necrosis factor and interferon resulted in a synergistic antiproliferative effect in vitro.' - :I In recent studies, a synergistic effect of tumour necrosis factor-a with interferons has been demonstrated both in vitro'> and in vivd against human pancreatic cancer cell lines. Tumour necrosis factor-B (TNF-~), which is also known as lymphotoxin, is produced by lymphocytes and has been shown to have a cytotoxic activity against some tumour cells in vitro. U.7 Few authors, however, have investigated the anti-proliferative effect of TNF-~ alone, or in combination with interferons, against pancreatic cancer cells. In the present study, the efficacy of natural human TNF -~ alone or in combination with natural human interferons against human pancreatic cancer cells in vitro was investigated.
CELL LINES The following human pancreatic cancer cell lines were used: PANC-l; MIA PaCa-2; and BxPC-3. These cell lines were obtained from the American Type Culture Collection (Rockville, USA) and were grown as monolayer cultures in 5% carbon dioxide/95% air at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated foetal calf serum, 100 IU/ml penicillin G and 100 IU/ml streptomycin sulfphate (Gibco, USA). Cells were maintained in logarithmic growth by replating, and cells in the exponential growth phase were harvested and used to study the effects of the cytokines.
REAGENTS Natural human TNF-~ and interferon-a were produced by a BALL-l cell line sensitized with haemagglutinating virus of Japan and were purified to over 99% homogeneity." Interferon-y was produced by lipopolysaccharide-stimulated human myel 0monocytic HBL-38 cells and was purified to over 99% homogeneity." The activity ofTNF~ was determined on the basis of its cytolytic activity against actinomycin D-treated L929 mouse fibroblasts using a method developed for tumour necrosis factor or lymphotoxin by Eifel et al." and was found to be 5 x 10" U/mg protein. The activities of interferon-a and interferon-y were determined using a cytopathic effect inhibition assay that
113 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fucbimoto, N Matsubara et al. Combination cytokine therapy
STATISTICAL ANALYSIS
employed human FL cells challenged with Sindbis virus. The results were standardized against the international reference preparations of human interferon-a (Ga23-901532) and interferon-y (Ga23-901-530), and the activities of natural human interferon-a and interferon-y were found to be 2 x lOA IU/mg protein and 2.1 x 10 7 IU/mg protein, respectively.
For evaluation of the interaction between TNF-~ and the interferons, median effect plot analysis was used." In brief, the data were plotted as y = log (F/FJ or its equivalence with respect to x e log (D), where Dis the dose, and F" and F" are the fractions of the system that were affected or unaffected, respectively, by the dose D. The pharmacological dose of each fraction was then calculated by linear regression. The combination index (CI) was then calculated using the following equation:
SENSITIVITY ASSAY Cells were seeded in triplicate into 96-well microtitre plates (Falcon Labware, USA) at 5 x 10 3 cells/well in 0.2 ml in the presence of various concentrations of the cytokines alone, or in combination. Medium alone was used for control cultures. The following combinations of cytokines were tested: TNF-~ with interferon-a (ratio 1 : 1) and TNF -~ with interferon-y (ratio 10 : 1). After incubation at 37°C in a humidified 5% carbon dioxide atmosphere for 96 h, the cells were stained with 0.05% crystal violet in 2% ethanol. The dye was eluted with Sorensons's buffer (6.1 ml 0.1 M disodium citrate, 3.9 ml 0.1 M HCI and 10 ml 95% ethanol; pH 5.2) and the absorbance at 540 nm was determined with a FT-2 plate analyser (Toyo Sokki Co., Japan). The number of viable cells was estimated from the absorbance, based on the correlation between the amount of dye eluted and the number of cells. Cell growth was calculated according to the following formula:
percentage growth
D, CI =
D2
+ D
X, l
D, D,
+ D
X •2
D.",Dx,'
where D 1 and D,,2 represent the corresponding doses for a given degree of effect, and D"',2 is the combined dose having the same degree of effect that can be separated into D, and D, by their known ratio. The interaction of two agents is considered to be synergistic when the CI value is less than 1, whereas for CI values equal to 1 the interaction is additive and for those greater than 1 it is antagonistic. X
•
RESULTS EFFECTS OF SINGLE CYTOKINES The anti-proliferative effect of TNF-~ as a single agent (1 - 10 5 U/ml) is shown in Fig. l(a) for the three human pancreatic cancer cell lines used. All cell lines exhibited a growth reduction of less than 20% after treatment with 10 5 U/ml TNF-~ when compared to untreated control cells. The susceptibility of the same panel of tumour cells to interferon-a was also tested and values are shown in Fig. l(b). Cell lines MIA PaCa-2 and BxPC-3 showed marked growth inhibition of about 70% after treatment with 10 5 IU/ml interferon-a, whereas PANC-1 cells showed growth inhibitions of only 20%. The growth
(mean absorbance of the treated group) - - - - - - - - - x 100 (mean absorbance of the control group)
Each experiment was performed in triplicate and was repeated at least three times.
114 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et al. Combination cytokine therapy
FIGURE 1
Anti-proliferative effect of (a) 1 - 1()5 Ulml natural human tumour necrosis factor-~, (b) 1 -1()5IUlml interferon-a and (c) 1 -1(J1IUlml imerteron-» on human pancreatic cancer eel/lines PANC-1, MIA PaCa-2 and BxPC-3.
EFFECT OF TUMOUR NECROSIS FACTOR-13 PLUS INTERFERON-y
inhibition produced by interferon-yis shown in Fig. Hc). The cell line BxPC-3 was highly sensitive to interferon-y (about 80% growth inhibition at 10 4 IV/ml) but MIA PaCa-2 and PANC-1 cells showed only slight and no growth inhibition, respectively.
When TNF-~ was combined with interferony at a ratio of 10: 1, synergistic inhibition of the growth of PANC -1 [Figs. 4(a) and 5(a)] and MIA PaCa-2 [Figs 4(b) and 5(b)] cells was observed. By contrast, the interaction was not synergistic for BxPC-3 cells [Fig. 5(c)], which were highly sensitive to interferon-y alone as shown in Fig. 4 (c). A summary of these results is shown in Table 1.
EFFECT OF TUMOUR NECROSIS FACTOR-13 PLUS INTERFERON-a The effect of the combination of TNF -~ and interferon-a on human pancreatic cancer cell lines was assessed using a combination ofthe two cytokines in the ratio of 1 : 1 at various concentrations (Fig. 2) and the CI values were calculated (Fig. 3). A synergistic interaction was observed for PANC-1 cells as shown in Fig. 3(a), whereas the interaction was only additive for MIA PaCa-2 cells [Fig. 3(b)] and BxPC-3 cells [Fig. 3(c)].
DISCUSSION The present study demonstrated synergism between natural human TNF -~ and interferons in their inhibition of the growth of pancreatic cancer cell lines. It has previously been shown that TNF -~ is toxic to a variety of cells and is
115 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et al, Combination cytokine therapy
FIGURE 2 (a)
120
(b)
120
100
100
100
80
80
80
~ 60
60
60
40
40
20
20
e
o
40 • TNF-~ • IFN-a • TNF-~/IFN-a
20
o0
10
10 2 10 3 10 4 10 5 0 0
(c)
120
1 10 102 103 104 105 Concentration (IU/ml)
0 0
10
10 2 10 3 10 4 10 5
Anti-proliferative effect of a combinationof 1 - 1fJ5 Ulml natural human tumour necrosis factor-I), (TNF-I) and 1 -1fJ5IUlml interferon-a, (IFN-a, ratio 1: 1) compared with 1 -1fJ5 Ulml TNF-I) or 1-1fJ5IUlmIIFN-a alone on human pancreatic cancercel/lines (a) PANC1, (b) MIA PaCa-2 and (c) BxPC-3.
FIGURE 3
(a)
2
01
(b)
~
r-,
~
0 0.1
(c)
~ 0.3
0.5
0.7
0.9 0.1
0.3
0.5
0.7
0.9
0.1
0.3
0.5
0.7
0.9
Fa
Relationship between the combination index (CI) and the fraction of the system that was affected (Fa)' calculated using the median effect method, for the assessment of the antiproliferative effect of a combination of 1-1OS U/ml natural human tumour necrosis factor-13 and 1 - 1fJ5/Ulml natural human interferon-a (ratio 1: 1) on human pancreatic cancer cel/lines (a) PAN-C, (b) MIA PaCa-2 and (c) BxPC-3 (CI < 1, synergistic effect; CI 1, additive effect; and CI> 1, antagonistic effect).
=
116 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
FIGUAE4
o
0
o
10
10 2 10 3 10 4 10 5
0 0
1
10
10 210 310 410 5
'-'---------'-_-'---------'--_.L--L..-----L.
0
10
10 2 10 3 10 4 10 5
Concentration (IU/ml)
Anti-proliferative effect of a combination of 10- 10S Ulml natural human tumour necrosis factor-~ (TNF-~) and 1 - 1(J4 IU/ml natural human inteferon..ry (IFN..ry, ratio 10 : 1) compared with 10 - 10S Ulml TNF-~ or 1 - 1(J4 IUlmIIFN..ry alone on human pancreatic cancer cel/lines (a) PANC-1, (b) MIA PaCa-2 and (c) BxPC-3.
(a)
2
\
1, antagonistic effect).
117 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
TABLE 1 Summary of the sensitivity of human pancreatic cancer eel/lines to natural human tumour necrosis factor-~ (TNF-~) and/or interferon-a (IFN-a) or interteron-t (IFN""{ )
Sensitivity·
Cell line
TNF-~
IFN-a
IFN-y
PANC-1
MIA PaCa-2
++
+
BxPC-3
+
++
a _,
TNF-~/IFN-a
+ (synergistic) +++ (additive) ++ (additive)
TNF-~/IFN·y
+ (synergistic) ++ (synergistic) +++ (additive)
> 75% growth; +, 50 - 74% growth; ++, 25 - 49% growth; +++, < 24% growth.
when used in combination with other cytokines. These findings, therefore, suggest that all types of pancreatic carcinoma should possibly be treated by combination cytokine therapy. The mechanisms by which the cytokines exerted an anti-proliferative effect on the tumour cells and showed synergism in combination therapy are still unclear. It has been reported that TNF-~ and TNF-a act via the same receptor and that the upregulation of TNF-receptor expression by interferons may be one of the mechanisms explaining synergism between these cytokines." - Hl Other reports have suggested that the antiproliferative effects and synergism of these cytokines may be due to the arrest of tumour cells at a specific phase of the cell cycle. 20 . 2 ' The exact mechanisms of action of these cytokines on target cells are under investigation, including events at the receptor and post-receptor levels. The present study has shown that the sensitivity of human pancreatic cancer cell lines to TNF-~ or interferons as single agents
structurally and functionally related to TNFa.'·12-14 Interferons regulate the activities of a variety of cells and tissues, including both those of the immune system and target cells during the immune response and many ofthe activities of tumour necrosis factor and interferon overlap each other. The combination of tumour necrosis factor and interferon often has a synergistic effect' - and TNF-~ plus interferons also exhibited a synergistic effect on human pancreactic cancer cells in the present study. The sensitivity of the pancreatic cancer cell lines used in the present study to each cytokine as a single agent varied from 0% to about 80% growth inhibition, whereas the combination of TNF-~ and interferons produced a synergistic effect in three of the six experiments. It is interesting to note when the cancer cells were resistant to each cytokine alone a marked enhancement of growth inhibition and a synergistic effect were obtained when the corresponding cytokines were used in combination. On the other hand, when a single agent had a potent anti-tumour activity, there was no synergism exhibited OJ
118 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et at. Combination cytokine therapy
was varied. Enhancement of growth inhibition was observed for the combination of TNF-~ and interferons when human pancreatic cancer cells were tested, especially for cells resistant to each cytokine alone. It is concluded that combination therapy with cytokines may have some potential activity against all types of pancreatic cancer and may lead to an improvement in the treatment of human pancreatic cancer. To improve the prognosis of patients with pancreatic cancer, however, studies including other new modalities are
needed, e.g. adoptive immunotherapy" and hormonal therapy,'" followed by clinical trials of any promising combination.
ACKNOWLEDGEMENTS The study was supported by Hayashibara Biochemical Laboratories, who kindly supplied the natural human TNF-~ and natural human interferon-a and interferon-y. We also wish to thank Mrs K Nasu and Miss S Segawa for their technical assistance.
REFERENCES 1 Lee SH, Aggarwal BB, Rinderknecht E, et a1: The synergistic anti-proliferative effect of y-interferon and human lymphotoxin. J Immuno11984; 133: 1083 - 1086. 2 Feng G, Gray PW, Shepard, et a1: Antiproliferative activity of a hybrid protein between interferon-y and tumor necrosis factor-B. Science 1988; 241: 1501 - 1503. 3 Naomoto Y, Tanaka N, Fuchimoto S, et a1: In vitro synergistic effects of natural human tumor necrosis factor and natural human interferon-a. [pii J Cancer Res 1987; 78: 87 - 92. 4 Schmeigel WH, Caesar J, Kalthoff H, et a1: Antiproliferative effects exerted by recombinant human tumor necrosis factora (TNF-a) and interferon-y (IFN-y) on human pancreatic tumor cell lines. Pancreas 1988; 3: 180 - 188. 5 Matsubara N, Fuchimoto S, Orita K: Antiproliferative effects of natural human tumor necrosis factor-a, interferon-a, and interferon-yon human pancreatic carcinoma cell lines. Int J Pancreato11991; 8: 235 - 243. 6 Kuhl JS, Klapdor R, BahIa M, et a1: Cyto-
7
8
9
10
11
kines and pancreatic cancer: sensitivity of xenotransplants of pancreatic carcinomas to rIFN-y and rTNF-a in nude mice. Int J Pancreato11989; 4: 303 - 319. Gray PW, Aggarwal BB, Benton CV, et a1: Cloning and expression of eDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature 1984; 312: 721 - 724. Fukada S, Ando S, Sanou 0, et a1: Simultaneous production of natural human tumor necrosis factor-a, -~ and interferona from BALL-l cells stimulated by HVJ. Lymphokine Res 1988; 7: 175 - 185. Ando S, Ohta T, Tanimoto T, et a1: Natural human interferon-y derived from lipopolysaccharide-stimulated human myelomonocytic HBL-38 cells. [pn J Cancer Res 1988; 79: 757 - 765. Eifel PJ, Walker SH, Lucas ZJ: Standardization of a sensitive and rapid assay for lymphotoxin. Celllmmunol1975; 15: 208 - 221. Chou TC, Talalay P: Quantitative analysis of dose - effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regu1 1984; 21:
119 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
27 - 48. 12 Shalaby MR, Aggarwal BB, Rinderknecht E, et aJ: Activation of human polymorphnuclear neutrophil functions by interferon-y and tumor necrosis factors. J ImmunoJ1985; 135: 2069 - 2073. 13 Kaushanski K, Broudy VC, Harlan JM, et aJ: Tumor necrosis factor-a and tumor necrosis factor-B (lymphotoxin) stimulate the production of granulocyte macrophage colony-stimulating factor, macrophage colony-stimulating factor, and IL-l in vivo. J ImmunoJ1988; 141: 34103415. 14 Zola H, Nikoloutsopoulos A: Effect of recombinant human tumor necrosis factor~ (TNF -~) on activiation, proliferation and differentiation of human beta lymphocytes. ImmunoJogy 1989; 67: 231 - 236. 15 Stauber GB, Aggarwal BB:Characterization and affinity cross-linking ofreceptors for human recombinant lymphotoxin (tumor necrosis factor-B) on a human histiocytic lymphoma cell line, U-937. J BioJ Chem 1989; 264: 3573 - 3576. 16 Hass PE, Hotchkiss A, Mohler M, et aJ: Characterization of specific high affinity receptors for human tumor necrosis factor on mouse fibroblasts. J BioJ Chern 1985; 260: 12214 - 12218. 17 Aggarwal BB, Eessalu TE, Has PE: Characterization of receptors for human tumour necrosis factor and their regulation by '1interferon. Nature 1985; 318: 665 -667. 18 Tsujimoto M, Yip YK, Vilcek ]: Interferon'1 enhances expression of cellular receptors for tumor necrosis factor. J ImrnunoJ
1986; 136: 2441 - 2444. 19 Ruggiero V, Tavernier J, Fiers W, et a1: Induction of the synthesis of tumor necrosis factor receptors by interferon-y. J ImrnunoJ1986; 136: 2445 - 2450. 20 Darzynkiewics Z, Williamson B, Carswell EA, et a1: Cell cycle-specific effects of tumor necrosis factor. Cancer Res 1984; 44: 83 - 90. 21 Muro M, Naomoto Y, Orita K: Mechanism ofthe combined antitumor effect of natural human tumor necrosis factor-a and natural human interferon-a on cell cycle progression. Jpn J Cancer Res 1991; 82: 118 -126. 22 MarincolaFM, DaPozzo LF, Drucker BJ, et aJ: Adoptive immunotherapy of human pancreatic cancer with lymphokineactivated killer cells and interleukin-2 in a nude mouse model. Surgery 1990; 108: 919 - 929. 23 Szende B, Srkalovic G, Groot K, et al: Regression of nitrosamine-induced pancreatic cancers in hamsters treated with luteinizing hormone-releasing hormone antagonists or agonists. Cancer Res 1990; 50: 3716 - 3721. T Watanabe, S Fuchimoto, N Matsubara, H Iwagaki and K Orita Anti-proliferative Effect on Human Pancreatic Cancer Cells of Natural Human Tumour Necrosis Factor-f Combined with Natural Human Interferon-a. or Interferon-y The Journal of International Medical Research 1992; 20: 112 -120
Received for publication 20 January 1992 Accepted 23 January 1992 © Copyright 1992 Cambridge Medical Publications Ltd
Address for correspondence DR T WATANABE First Department of Surgery, Okayarna University Medical School, 2-5-1 Shikata-cho, Dkayama 700, Japan.
120 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
WATANABE,
S
FUCHIMOTO,
N
MATSUBARA,
H
IWAGAKI AND
K
ORITA
First Department of Surgery, Okayama University Medical School, Okayama, Japan
The anti-proliferative effects of natural cytokines, human tumour necrosis factor-B. natural human interferon-a and natural human interferon-y, on three human pancreatic cancer cell lines (PANC-l, MIA PaCa-2 and BxPC-3) were investigated in vitro. The anti-proliferative effect was determined using the dye uptake method and analysed for synergism by the median effect principle. Tumour necrosis factor-B. as a single agent, had little anti-proliferative effect on any of the three cell lines, whereas interferon-a and interferon-y exhibited a strong anti-proliferative effect against two cell lines (MIA PaCa-2 and BxPC-3) and one cell line (BxPC-3), respectively. When tumour necrosis factor-B and interferon-a were administered together (ratio 1: 1), a synergistic effect was observed against PANC-l cells. The combination of tumour necrosis factor-f and interferon-y (ratio 10: 1) was synergisic against both PANC-l and MIA PaCa-2 cells. A synergistic anti-proliferative effect oftumour necrosis factor-B and interferons was, therefore, observed even for cell lines that showed little biological response to
112 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et at. Combination cytokine therapy
each cytokine alone. The data suggest that some future improvement in the treatment of pancreatic cancer may be obtained by using combination cytokine therapy. KEY WORDS:
TUMOUR NECROSIS
FACTOR-P;
INTERFERON-a;
INTERFERON-y; CYTOKINES; HUMAN PANCREATIC CANCER; IMMUNOTHERAPY
MATERIALS AND METHODS
INTRODUCTION Pancreatic cancer has one of the worst prognoses among all the malignancies and, despite many efforts to improve the therapy available for this tumour, no satisfactory progress has been achieved. Effective adjuvant therapy, such as radiotherapy, chemotherapy, immunotherapy, hyperthermia or hormonal therapy, is required to supplement surgery. The efficacy of several biological response modifiers for pancreatic cancer has been investigated, but most have produced little response. It has been reported that the treatment of various tumour cell lines with tumour necrosis factor and interferon resulted in a synergistic antiproliferative effect in vitro.' - :I In recent studies, a synergistic effect of tumour necrosis factor-a with interferons has been demonstrated both in vitro'> and in vivd against human pancreatic cancer cell lines. Tumour necrosis factor-B (TNF-~), which is also known as lymphotoxin, is produced by lymphocytes and has been shown to have a cytotoxic activity against some tumour cells in vitro. U.7 Few authors, however, have investigated the anti-proliferative effect of TNF-~ alone, or in combination with interferons, against pancreatic cancer cells. In the present study, the efficacy of natural human TNF -~ alone or in combination with natural human interferons against human pancreatic cancer cells in vitro was investigated.
CELL LINES The following human pancreatic cancer cell lines were used: PANC-l; MIA PaCa-2; and BxPC-3. These cell lines were obtained from the American Type Culture Collection (Rockville, USA) and were grown as monolayer cultures in 5% carbon dioxide/95% air at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated foetal calf serum, 100 IU/ml penicillin G and 100 IU/ml streptomycin sulfphate (Gibco, USA). Cells were maintained in logarithmic growth by replating, and cells in the exponential growth phase were harvested and used to study the effects of the cytokines.
REAGENTS Natural human TNF-~ and interferon-a were produced by a BALL-l cell line sensitized with haemagglutinating virus of Japan and were purified to over 99% homogeneity." Interferon-y was produced by lipopolysaccharide-stimulated human myel 0monocytic HBL-38 cells and was purified to over 99% homogeneity." The activity ofTNF~ was determined on the basis of its cytolytic activity against actinomycin D-treated L929 mouse fibroblasts using a method developed for tumour necrosis factor or lymphotoxin by Eifel et al." and was found to be 5 x 10" U/mg protein. The activities of interferon-a and interferon-y were determined using a cytopathic effect inhibition assay that
113 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fucbimoto, N Matsubara et al. Combination cytokine therapy
STATISTICAL ANALYSIS
employed human FL cells challenged with Sindbis virus. The results were standardized against the international reference preparations of human interferon-a (Ga23-901532) and interferon-y (Ga23-901-530), and the activities of natural human interferon-a and interferon-y were found to be 2 x lOA IU/mg protein and 2.1 x 10 7 IU/mg protein, respectively.
For evaluation of the interaction between TNF-~ and the interferons, median effect plot analysis was used." In brief, the data were plotted as y = log (F/FJ or its equivalence with respect to x e log (D), where Dis the dose, and F" and F" are the fractions of the system that were affected or unaffected, respectively, by the dose D. The pharmacological dose of each fraction was then calculated by linear regression. The combination index (CI) was then calculated using the following equation:
SENSITIVITY ASSAY Cells were seeded in triplicate into 96-well microtitre plates (Falcon Labware, USA) at 5 x 10 3 cells/well in 0.2 ml in the presence of various concentrations of the cytokines alone, or in combination. Medium alone was used for control cultures. The following combinations of cytokines were tested: TNF-~ with interferon-a (ratio 1 : 1) and TNF -~ with interferon-y (ratio 10 : 1). After incubation at 37°C in a humidified 5% carbon dioxide atmosphere for 96 h, the cells were stained with 0.05% crystal violet in 2% ethanol. The dye was eluted with Sorensons's buffer (6.1 ml 0.1 M disodium citrate, 3.9 ml 0.1 M HCI and 10 ml 95% ethanol; pH 5.2) and the absorbance at 540 nm was determined with a FT-2 plate analyser (Toyo Sokki Co., Japan). The number of viable cells was estimated from the absorbance, based on the correlation between the amount of dye eluted and the number of cells. Cell growth was calculated according to the following formula:
percentage growth
D, CI =
D2
+ D
X, l
D, D,
+ D
X •2
D.",Dx,'
where D 1 and D,,2 represent the corresponding doses for a given degree of effect, and D"',2 is the combined dose having the same degree of effect that can be separated into D, and D, by their known ratio. The interaction of two agents is considered to be synergistic when the CI value is less than 1, whereas for CI values equal to 1 the interaction is additive and for those greater than 1 it is antagonistic. X
•
RESULTS EFFECTS OF SINGLE CYTOKINES The anti-proliferative effect of TNF-~ as a single agent (1 - 10 5 U/ml) is shown in Fig. l(a) for the three human pancreatic cancer cell lines used. All cell lines exhibited a growth reduction of less than 20% after treatment with 10 5 U/ml TNF-~ when compared to untreated control cells. The susceptibility of the same panel of tumour cells to interferon-a was also tested and values are shown in Fig. l(b). Cell lines MIA PaCa-2 and BxPC-3 showed marked growth inhibition of about 70% after treatment with 10 5 IU/ml interferon-a, whereas PANC-1 cells showed growth inhibitions of only 20%. The growth
(mean absorbance of the treated group) - - - - - - - - - x 100 (mean absorbance of the control group)
Each experiment was performed in triplicate and was repeated at least three times.
114 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et al. Combination cytokine therapy
FIGURE 1
Anti-proliferative effect of (a) 1 - 1()5 Ulml natural human tumour necrosis factor-~, (b) 1 -1()5IUlml interferon-a and (c) 1 -1(J1IUlml imerteron-» on human pancreatic cancer eel/lines PANC-1, MIA PaCa-2 and BxPC-3.
EFFECT OF TUMOUR NECROSIS FACTOR-13 PLUS INTERFERON-y
inhibition produced by interferon-yis shown in Fig. Hc). The cell line BxPC-3 was highly sensitive to interferon-y (about 80% growth inhibition at 10 4 IV/ml) but MIA PaCa-2 and PANC-1 cells showed only slight and no growth inhibition, respectively.
When TNF-~ was combined with interferony at a ratio of 10: 1, synergistic inhibition of the growth of PANC -1 [Figs. 4(a) and 5(a)] and MIA PaCa-2 [Figs 4(b) and 5(b)] cells was observed. By contrast, the interaction was not synergistic for BxPC-3 cells [Fig. 5(c)], which were highly sensitive to interferon-y alone as shown in Fig. 4 (c). A summary of these results is shown in Table 1.
EFFECT OF TUMOUR NECROSIS FACTOR-13 PLUS INTERFERON-a The effect of the combination of TNF -~ and interferon-a on human pancreatic cancer cell lines was assessed using a combination ofthe two cytokines in the ratio of 1 : 1 at various concentrations (Fig. 2) and the CI values were calculated (Fig. 3). A synergistic interaction was observed for PANC-1 cells as shown in Fig. 3(a), whereas the interaction was only additive for MIA PaCa-2 cells [Fig. 3(b)] and BxPC-3 cells [Fig. 3(c)].
DISCUSSION The present study demonstrated synergism between natural human TNF -~ and interferons in their inhibition of the growth of pancreatic cancer cell lines. It has previously been shown that TNF -~ is toxic to a variety of cells and is
115 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et al, Combination cytokine therapy
FIGURE 2 (a)
120
(b)
120
100
100
100
80
80
80
~ 60
60
60
40
40
20
20
e
o
40 • TNF-~ • IFN-a • TNF-~/IFN-a
20
o0
10
10 2 10 3 10 4 10 5 0 0
(c)
120
1 10 102 103 104 105 Concentration (IU/ml)
0 0
10
10 2 10 3 10 4 10 5
Anti-proliferative effect of a combinationof 1 - 1fJ5 Ulml natural human tumour necrosis factor-I), (TNF-I) and 1 -1fJ5IUlml interferon-a, (IFN-a, ratio 1: 1) compared with 1 -1fJ5 Ulml TNF-I) or 1-1fJ5IUlmIIFN-a alone on human pancreatic cancercel/lines (a) PANC1, (b) MIA PaCa-2 and (c) BxPC-3.
FIGURE 3
(a)
2
01
(b)
~
r-,
~
0 0.1
(c)
~ 0.3
0.5
0.7
0.9 0.1
0.3
0.5
0.7
0.9
0.1
0.3
0.5
0.7
0.9
Fa
Relationship between the combination index (CI) and the fraction of the system that was affected (Fa)' calculated using the median effect method, for the assessment of the antiproliferative effect of a combination of 1-1OS U/ml natural human tumour necrosis factor-13 and 1 - 1fJ5/Ulml natural human interferon-a (ratio 1: 1) on human pancreatic cancer cel/lines (a) PAN-C, (b) MIA PaCa-2 and (c) BxPC-3 (CI < 1, synergistic effect; CI 1, additive effect; and CI> 1, antagonistic effect).
=
116 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
FIGUAE4
o
0
o
10
10 2 10 3 10 4 10 5
0 0
1
10
10 210 310 410 5
'-'---------'-_-'---------'--_.L--L..-----L.
0
10
10 2 10 3 10 4 10 5
Concentration (IU/ml)
Anti-proliferative effect of a combination of 10- 10S Ulml natural human tumour necrosis factor-~ (TNF-~) and 1 - 1(J4 IU/ml natural human inteferon..ry (IFN..ry, ratio 10 : 1) compared with 10 - 10S Ulml TNF-~ or 1 - 1(J4 IUlmIIFN..ry alone on human pancreatic cancer cel/lines (a) PANC-1, (b) MIA PaCa-2 and (c) BxPC-3.
(a)
2
\
1, antagonistic effect).
117 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
TABLE 1 Summary of the sensitivity of human pancreatic cancer eel/lines to natural human tumour necrosis factor-~ (TNF-~) and/or interferon-a (IFN-a) or interteron-t (IFN""{ )
Sensitivity·
Cell line
TNF-~
IFN-a
IFN-y
PANC-1
MIA PaCa-2
++
+
BxPC-3
+
++
a _,
TNF-~/IFN-a
+ (synergistic) +++ (additive) ++ (additive)
TNF-~/IFN·y
+ (synergistic) ++ (synergistic) +++ (additive)
> 75% growth; +, 50 - 74% growth; ++, 25 - 49% growth; +++, < 24% growth.
when used in combination with other cytokines. These findings, therefore, suggest that all types of pancreatic carcinoma should possibly be treated by combination cytokine therapy. The mechanisms by which the cytokines exerted an anti-proliferative effect on the tumour cells and showed synergism in combination therapy are still unclear. It has been reported that TNF-~ and TNF-a act via the same receptor and that the upregulation of TNF-receptor expression by interferons may be one of the mechanisms explaining synergism between these cytokines." - Hl Other reports have suggested that the antiproliferative effects and synergism of these cytokines may be due to the arrest of tumour cells at a specific phase of the cell cycle. 20 . 2 ' The exact mechanisms of action of these cytokines on target cells are under investigation, including events at the receptor and post-receptor levels. The present study has shown that the sensitivity of human pancreatic cancer cell lines to TNF-~ or interferons as single agents
structurally and functionally related to TNFa.'·12-14 Interferons regulate the activities of a variety of cells and tissues, including both those of the immune system and target cells during the immune response and many ofthe activities of tumour necrosis factor and interferon overlap each other. The combination of tumour necrosis factor and interferon often has a synergistic effect' - and TNF-~ plus interferons also exhibited a synergistic effect on human pancreactic cancer cells in the present study. The sensitivity of the pancreatic cancer cell lines used in the present study to each cytokine as a single agent varied from 0% to about 80% growth inhibition, whereas the combination of TNF-~ and interferons produced a synergistic effect in three of the six experiments. It is interesting to note when the cancer cells were resistant to each cytokine alone a marked enhancement of growth inhibition and a synergistic effect were obtained when the corresponding cytokines were used in combination. On the other hand, when a single agent had a potent anti-tumour activity, there was no synergism exhibited OJ
118 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
T Watanabe, S Fuchimoto, N Matsubara et at. Combination cytokine therapy
was varied. Enhancement of growth inhibition was observed for the combination of TNF-~ and interferons when human pancreatic cancer cells were tested, especially for cells resistant to each cytokine alone. It is concluded that combination therapy with cytokines may have some potential activity against all types of pancreatic cancer and may lead to an improvement in the treatment of human pancreatic cancer. To improve the prognosis of patients with pancreatic cancer, however, studies including other new modalities are
needed, e.g. adoptive immunotherapy" and hormonal therapy,'" followed by clinical trials of any promising combination.
ACKNOWLEDGEMENTS The study was supported by Hayashibara Biochemical Laboratories, who kindly supplied the natural human TNF-~ and natural human interferon-a and interferon-y. We also wish to thank Mrs K Nasu and Miss S Segawa for their technical assistance.
REFERENCES 1 Lee SH, Aggarwal BB, Rinderknecht E, et a1: The synergistic anti-proliferative effect of y-interferon and human lymphotoxin. J Immuno11984; 133: 1083 - 1086. 2 Feng G, Gray PW, Shepard, et a1: Antiproliferative activity of a hybrid protein between interferon-y and tumor necrosis factor-B. Science 1988; 241: 1501 - 1503. 3 Naomoto Y, Tanaka N, Fuchimoto S, et a1: In vitro synergistic effects of natural human tumor necrosis factor and natural human interferon-a. [pii J Cancer Res 1987; 78: 87 - 92. 4 Schmeigel WH, Caesar J, Kalthoff H, et a1: Antiproliferative effects exerted by recombinant human tumor necrosis factora (TNF-a) and interferon-y (IFN-y) on human pancreatic tumor cell lines. Pancreas 1988; 3: 180 - 188. 5 Matsubara N, Fuchimoto S, Orita K: Antiproliferative effects of natural human tumor necrosis factor-a, interferon-a, and interferon-yon human pancreatic carcinoma cell lines. Int J Pancreato11991; 8: 235 - 243. 6 Kuhl JS, Klapdor R, BahIa M, et a1: Cyto-
7
8
9
10
11
kines and pancreatic cancer: sensitivity of xenotransplants of pancreatic carcinomas to rIFN-y and rTNF-a in nude mice. Int J Pancreato11989; 4: 303 - 319. Gray PW, Aggarwal BB, Benton CV, et a1: Cloning and expression of eDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature 1984; 312: 721 - 724. Fukada S, Ando S, Sanou 0, et a1: Simultaneous production of natural human tumor necrosis factor-a, -~ and interferona from BALL-l cells stimulated by HVJ. Lymphokine Res 1988; 7: 175 - 185. Ando S, Ohta T, Tanimoto T, et a1: Natural human interferon-y derived from lipopolysaccharide-stimulated human myelomonocytic HBL-38 cells. [pn J Cancer Res 1988; 79: 757 - 765. Eifel PJ, Walker SH, Lucas ZJ: Standardization of a sensitive and rapid assay for lymphotoxin. Celllmmunol1975; 15: 208 - 221. Chou TC, Talalay P: Quantitative analysis of dose - effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regu1 1984; 21:
119 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
27 - 48. 12 Shalaby MR, Aggarwal BB, Rinderknecht E, et aJ: Activation of human polymorphnuclear neutrophil functions by interferon-y and tumor necrosis factors. J ImmunoJ1985; 135: 2069 - 2073. 13 Kaushanski K, Broudy VC, Harlan JM, et aJ: Tumor necrosis factor-a and tumor necrosis factor-B (lymphotoxin) stimulate the production of granulocyte macrophage colony-stimulating factor, macrophage colony-stimulating factor, and IL-l in vivo. J ImmunoJ1988; 141: 34103415. 14 Zola H, Nikoloutsopoulos A: Effect of recombinant human tumor necrosis factor~ (TNF -~) on activiation, proliferation and differentiation of human beta lymphocytes. ImmunoJogy 1989; 67: 231 - 236. 15 Stauber GB, Aggarwal BB:Characterization and affinity cross-linking ofreceptors for human recombinant lymphotoxin (tumor necrosis factor-B) on a human histiocytic lymphoma cell line, U-937. J BioJ Chem 1989; 264: 3573 - 3576. 16 Hass PE, Hotchkiss A, Mohler M, et aJ: Characterization of specific high affinity receptors for human tumor necrosis factor on mouse fibroblasts. J BioJ Chern 1985; 260: 12214 - 12218. 17 Aggarwal BB, Eessalu TE, Has PE: Characterization of receptors for human tumour necrosis factor and their regulation by '1interferon. Nature 1985; 318: 665 -667. 18 Tsujimoto M, Yip YK, Vilcek ]: Interferon'1 enhances expression of cellular receptors for tumor necrosis factor. J ImrnunoJ
1986; 136: 2441 - 2444. 19 Ruggiero V, Tavernier J, Fiers W, et a1: Induction of the synthesis of tumor necrosis factor receptors by interferon-y. J ImrnunoJ1986; 136: 2445 - 2450. 20 Darzynkiewics Z, Williamson B, Carswell EA, et a1: Cell cycle-specific effects of tumor necrosis factor. Cancer Res 1984; 44: 83 - 90. 21 Muro M, Naomoto Y, Orita K: Mechanism ofthe combined antitumor effect of natural human tumor necrosis factor-a and natural human interferon-a on cell cycle progression. Jpn J Cancer Res 1991; 82: 118 -126. 22 MarincolaFM, DaPozzo LF, Drucker BJ, et aJ: Adoptive immunotherapy of human pancreatic cancer with lymphokineactivated killer cells and interleukin-2 in a nude mouse model. Surgery 1990; 108: 919 - 929. 23 Szende B, Srkalovic G, Groot K, et al: Regression of nitrosamine-induced pancreatic cancers in hamsters treated with luteinizing hormone-releasing hormone antagonists or agonists. Cancer Res 1990; 50: 3716 - 3721. T Watanabe, S Fuchimoto, N Matsubara, H Iwagaki and K Orita Anti-proliferative Effect on Human Pancreatic Cancer Cells of Natural Human Tumour Necrosis Factor-f Combined with Natural Human Interferon-a. or Interferon-y The Journal of International Medical Research 1992; 20: 112 -120
Received for publication 20 January 1992 Accepted 23 January 1992 © Copyright 1992 Cambridge Medical Publications Ltd
Address for correspondence DR T WATANABE First Department of Surgery, Okayarna University Medical School, 2-5-1 Shikata-cho, Dkayama 700, Japan.
120 Downloaded from imr.sagepub.com at KAI NAN UNIV on February 12, 2015
