Radiol Oncol 1992; 26: 223-7.
Serum TNF-a levels in melanoma-bearing and healthy mice
Breda Kus1' 3, Gregor Sersa2, Srdjan Novakovic2 and Anton Stale1
1	LEK Pharmaceutical and Chemical Company, d. d., Research and Development.
2	Institute of Oncology, Department of Tumor Biology and Biotherapy, Ljubljana.
3 Institute of Chemistry, Ljubljana.
To determine differences in tumor necrosis factor-a (TNF-a) persistency after TNF-a therapy, serum levels were monitored in tumor bearing and healthy mice. Melanoma B-16-bearing mice were treated peritumorally and healthy mice subcutaneously with recombinant human TNF-a, lacking 1-3 amino acids from N-terminal part (TNF-aNv3). In healthy mice the peak TNF-a serum levels were detected one hour after application, while in tumor-bearing mice two hours after application. Also TNF-a elimination from serum of tumor-bearing mice was slower compared to healthy mice. The data suggest that TNF-aNv3 pharmacokinetic mechanisms in tumor-bearing and healthy mice are different.
Key words: tumor necrosis factor; pharmacokinetics; melanoma, experimental; mice
Introduction
Tumor necrosis factor-a (TNF-a) is a monocyte/ macrophage derived protein, originally identified by its ability to induce haemorrhagic necrosis of some tumors in vivo1 and cytotoxicity against certain tumor cells in vitro.2 The results of clinical trials with recombinant human TNF-a indicate that the side effects of TNF-a treatment can be severe.3 4 For rational dose regimen, which could diminish the side effects, it is important to know how TNF-a is absorbed from the site of application, distributed through the body, and its activity preserved.
Correspondence to: Breda Kus MSc, LEK Pharmaceutical and Chemical Company, d. d., Research and Development, 61007 Ljubljana, Celovška 135, Slovenia.
Despite the fact that TNF-a activity is at least partially species specific, animal models were often used to study pharmacokinetics of human TNF-a. However in all previous pharmacokinetic studies in mice,5 rats,67 rabbits and monkeys,8 human recombinant TNF-a was applied to healthy animals.
Therefore the aim of our study was to compare absorption and distribution of human TNF-a in tumor-bearing and healthy mice. Besides antitumor effect, serum TNF-a levels were measured after local TNF-a application. Human recombinant TNF-a, lacking 1-3 amino acids from N-terminal part (TNF-aNv3), was applied peritumorally in melanoma B-16 bearing mice and subcutaneously in healthy mice.
UDC: 616-066, 81-085-092,9
224
Kus Bet al.
Materials and methods
Animals
Inbred C57B1/6 mice were purchased from Ru-djer Boskovic Institute, Zagreb, Croatia. Animals were maintained in constant room temperature (24° C) at a natural day/night cycle in a conventional mouse colony. Mice in good condition, without signs of fungal or other infections, 8-10 weeks old, were used in the experiments.
Tumors
Melanoma B-16 was maintained in C57Bl/6 mice by serial transplantation. Tumor cells from the fourth isotransplantat generation were prepared for the described experiments by gentle mechanical disagregation. Solid subcutaneous tumors, dorsolaterally in animals, were initiated by injection of 5 x 105 viable melanoma cells. The viability of the cells was determined by trypan blue dye exclusion test.9
Tumor necrosis factor
Recombinant human tumor necrosis factor-a analog, lacking 1-3 amina acids from N-terminal end (TNF-aNv3), was used (ZIMET, Jena, Germany). Specific activity was 2.2 x 107 U/mg, tested on L929 cells in the presence of actino-mycin-D. Other properties of TNF-aNv3 were described before. 10 TNF-aNv3 was diluted in PBS (pH 7.4) before use.
Treatment
TNF-aNv3 was administered subcutaneously in tumor free and peritumorally in tumor-bearing mice.9 The treatment was started when the tumors reached 30-40 mm3 in volume. TNF-aNv3 application dose was 2 x 105 U per animal, which gave 9 x 103 ± 1039 U/g of animal.
Tumor measurement
Tumor growth was followed by measuring three mutually orthogonal tumor diameters with a Vernier calliper. Tumor volumes were calculated by formula rr/6 • a ■ b ■ c (a, b, c are tumor diameters). Fram the measurements, arithmetic means (AM) and standard errors of the means (SE) were calculated for each experimental group with minimum of 10 mice. Growth delay of the tumors was calculated from individual tumor growth, by determining the time required for tumor to reach 150 mm3 of volume, and subtracting the relevant time in the control
9 11
group."
Blood sample collection and handling
One day before TNF-aNv3 application blood was collected from all animals in the experiment for monitoring the serum TNF-a level. After the application of TNF-aNv3 blood samples were taken from four animals at the times indicated in the figures. Blood was collected from the orbital sinus of the animals, immediately centrifuged (3000 rpm/min) for 10 min at 4° C and sterilised by filtration (0.22 ^m cellulose acetate filter, Costar). Samples were stored at -70° C for TNF-a assay.
Cells and culture conditions
L929 murine transformed fibroblasts were obtained from Istituto Zooprofilattico Sperimen-tale, Brescia, Italy. Cells were cultured in Eagle's minimal essential medium (EMEM) with 10% fetal calf serum (FCS). Incubation was carried out at 37° C in a humidified 5% CO2 incubator.
Determination of TNF-a serum concentrations
a) L929 bioassay. Cytotoxicity of TNF-a was determined on L929 cells12 in the presence of actinomycin D. The viable cells density was determined 20 h after TNF-aNv3 treatment.
Serum TNFa levels in melanoma-bearing and healthy mice
225
Culture medium was removed and the cells, fixed with glutaraldehyde, were stained with crystal violet (0.5 % in 20 % methanol). Bound dye was eluted with 0.1 ml of 1% sodium dodecyl sulfate. The optical density was measured at 540 nm using microplate photometer CLS962 (Cambridge Life Sciences pic). Cytotoxic activity expressed as TNF-a U/ml was defined as the reciprocal of the dilution which gives 50% cell killing. The detection limit of the assay was 40 U/ml of serum. Cytotoxic activity for one sample was calculated as AM of six independent measurements. AM of three TNF-aNv3 activities in sera prepared from three mice in test group, are presented in figures.
b) TNF-a ELISA. Immunologically active TNF-a was determined by a "sandwich" enzyme immunosorbent assay (ELISA) specific for human TNF-a (Du Pont). The detection limit was 100 pg/ml of serum. TNF-aNv3 value for one sample was calculated as AM of two independent measurements. AM of three TNF-aNv3 values in sera prepared from three mice in test group, are presented in the results.
Statistical evaluation
The results were evaluated by Student's unpaired probability test and correlation coefficient. All statistical procedures were computed using CSS Programme Stat Soft. Levels of less than 0.05 were taken as indicating significant differences.
Results
Antitumor effect of TNF-aNv3
Melanoma B-16-bearing mice were treated pe-ritumorally with 2 x 105 U TNF-aNv3, when tumors reached average tumor volume 34.4 ± 3.9 mm3. Tumor growth delay after TNF-aNv3 treatment (3.8 ± 0.4 days) was significant (P < 0.05), compared to tumor growth in control animals, which were treated with physiological saline (Figure 1). No side effects were observed after peritumoral treatment with TNF-aNv3.
1000
G	8	10
DAYS AFTER TREATMENT
Figure l. Antitumor effect of TNF-aNv3 on subcutaneous melanoma B-16 after peritumoral (•) treatment with TNF-aNv3 (2 x 105 U/mouse); controls (•).
Endogenous TNF-a levels in mice before application of TNF-aNv3
Endogenous serum TNF-a levels were measured with L929-cell assay method, in sera collected one day before the TNF-aNv3 application from tumor-bearing and healthy animals. In all serum samples tested (15 from healthy and 24 from tumor bearing animals), endogenous TNF-a levels were below the reliable detection limit (40 U/ml) of the method.
The same serum samples were tested also on ELISA to test the method for possible interference of other immunological substances. Results were negative for all serum samples. This confirms that no component in the murine serum, prepared as described, interferes with our ELISA assay system.
Serum levels of TNF-aNv3 after peritumoral application
Serum TNF-a concentrations in B-16-mela-noma bearing mice were compared to concentrations of TNF-a in healthy mice without tumors. Tumor-bearing mice were treated peritu-
226
Kus Bet al.
MINUTES AFTER TREATMENT
Figure 2. Mean ± SE serum levels time curves of cytotoxically active TNF-aNv3 after peritumoral application (2 x 105 U/mouse) in B-16-bearing (•) and healthy mice (•).
morally, while healthy mice subcutaneously with 2 x 105 U TNF-aNv3. In the same serum samples, TNF-a cytotoxic activity (Figure 2) and immunologically active TNF-a molecules (Figure 3) were determined. Delayed TNF-a serum peak levels were observed in tumor-bearing mice (120 min.) compared to healthy mice (60 min.). Also, TNF-a serum concentrations were higher in tumor-bearing mice. The highest TNF-a serum level in tumor-bearing mice was 3.6 x 103 U/ml (L929 assay) or 8 x 104 pg/ml (ELISA), and 1.6 x 103 U/ml or 6 x 103 pg/ml in healthy mice (P < 0.05 in both assays). High TNF-a serum levels persisted in tumor-bearing animals, compared to quick clearance from healthy mice. Ten hours after TNF-aNv3 application its level in tumor bearing animals was 2.8 x 102 U/ml or 7 x 103 pg/ml, while in healthy mice already after four hours dropped to 83 U/ml or 1 x 102 pg/ml.
Discussion
The in vivo antitumor effects of murine TNF-a13 and human TNF-a14 have been reported on
1000000
o u
o 100 200 300 400 500 600 MINUTES AFTER TREATMENT
Figure 3. Mean ± SE serum levels time curves of TNF-aNv3 antigen after peritumoral application (2 x 105 U/mouse) in B-16-bearing (•) and healthy mice (•).
different murine tumors transplanted in synge-nic mice. In our previous study antitumor effects of TNF-a analog TNF-aNv3 after peritumoral application was established on SAl sarcoma and B-16 melanoma tumors.911 In this study the potent antitiumor effect of TNF-aNv3 is reconfirmed on B-16 melanoma tumors.
The results of TNF-a serum level measurement show that TNF-aNv3 pharmacokinetic mechanisms in tumor-bearing and healthy mice are different. TNF-a concentration-time profiles, presented in Figures 2 and 3 indicate, that TNF-a in serum of B-16-bearing mice persists longer than in serum of healthy mice.
TNF-aNv3 was measured in sera cytotoxically and immunologically in tumor-bearing and healthy mice after peritumoral or subcutaneous application, respectively. Relationship between results obtained by the two assays shows almost perfect positive correlation (R = 0.88 for healthy and R = 0.98 for tumor-bearing mice). Besides, it was established that no component from sera interfered with ELISA assay used. Although TNF-a levels were often found to be increased in sera of cancer patients15 we didn't detect endogenous TNF-a before TNF-aNv3
Serum TNF-a levels in melanoma-bearing and healthy mice
227
application neither in sera from healthy nor in sera from tumor-bearing mice. Therefore we believe that all TNF-a activity detected in serum originates from TNF-aNv3 applied.
The influence of B-16 tumor on the persistence of TNF-a activity in serum is documented for the first time. If other tumors also affect the persistence of TNF-a serum concentration in a similar way, the pharmacokinetic parameters from healthy animals can not be extrapolated to tumor-bearing animals; consequently for the prediction of human therapy only pharma-cokinetic data obtained by tumor-bearing animal models can be used.
In conclusion, this study demonstrates that peritumoral treatment with TNF-a analog, lacking first three amino acids from N- terminal part, has antitumor effect on B-16 tumors. Comparison of TNF-a serum concentration-time profiles in tumor-bearing and healthy mice, indicate that the absorption, distribution and/or metabolism of TNF-Nv3 in B-16 tumor-bearing mice are significantly different from those in healthy mice. The mechanism for mentioned differences between tumor-bearing and healthy mice remains to be clarified in further studies.
Acknowledgement
The financial support by grant A185 of the Ministry for Science and Technology of Slovenia is gratefully acknowledged.
References
1.	Carswell EA, Old LJ, Kassel RL, Green S, Fiore N, Williamson B. An endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci USA 1975 ; 72: 3666-70.
2.	Helson L, Green S, Carswell EA, Old LJ. Effect on tumor necrosis factor on cultured human melanoma cells. Nature 1975; 258 : 731-2.
3.	Blick M, Sherwin SA, Rosenblum M, Gutterman J. Phase I study of recombinant tumor necrosis factor in cancer patients. Cancer Res 1987; 47: 2986-9.
4.	Kramer SM, Sherwin SA. Tumor necrosis factor: drug development perspective. Drug News Per-spect 1989; 2: 214-7.
5.	Noguchi K, Inagawa H, Tsuji Y, Morikawa A, Mizuno DI, Soma GI. Antitumor activity of a novel chimera tumor necrosis factor (TNF- STH) constructed by connecting rTNF-S with thymosin against murine syngenic tumors. J Immunother 1991; 10: 105-11.
6.	ZahnG, Greischel A. Pharmacokinetics of tumor necrosis factor alpha after intravenous administration in rats. Drug Res 1989; 39: 1180-2.
7.	Ferraiolo BL, McCabe J, Hollenbach S, Hultgren B, Pitti R, Wilking H. Pharmacokinetics of recombinant human tumor necrosis factor-a in rats. Drug Methab Dispos 1989; 17 : 369-72.
8.	Bocci V, Paccini A, Pessina P, Maioli E, Naldini A. Studies on tumor necrosis factor (TNF) I. Pharmacokinetics of human recombinant TNF in rabbits and monkeys after intravenous administration. Gen Pharmacol 1987; 18: 343-6.
9.	Sersa G, Novakovic S, Stalc A. Tumor mass is a critical factor in peritumoral treatment with tumor necrosis factor. Period Biol 1992; 94 : 35-40.
10.	Kus B, Stalc A. Biological activities of a derivative of human tumor necrosis factor-a lacking three N-terminal amina acids. Acta Pharm 1992; 42: 117-24.
11.	Sersa G, Novakovic S, Stalc A. Antitumor effect of recombinant human tumor necrosis factor-a analog combined with desmuramyl dipeptides LK-409 or LK-410 on sarcoma in mice. Mol Biother 1992; 4: in press.
12.	Matthews N, Neale ML. Cytotoxicity assays for tumor necrosis factor and lymphotoxin. In: Clemens MJ, Morris AG, Gearing AJH eds. Lym-phokines and Interferones Oxford: IRL Press, 1987: 221-5.
13.	Haranaka K, Satomi N, Sakurai A. Antitumor activity of murine tumor necrosis factor (TNF) against transplanted murine tumors and hetero-transplanted human tumors in nude mice. Int J Cancer 1984; 34: 263-7.
14.	Sohmura Y, Nakata K, Yoshida H, Kashimoto S, Matsui Y, Furuichi H. Recombinant human tumor necrosis factor - II. Antitumor effects on murine and human tumors transplanted in mice. Int J Immunopharmac 1986; 8: 357-68.
15.	Aderka D, Fisher S, Levo Y, Holtman H, Hahn T, Wallach D. Cachectin/tumor necrosis factor production by cancer patients. Lancet 1985 ; 2: 1190.