INFUSION OF TUMOR NECROSIS FACTOR (TNF) CAUSES AN INCREASE IN CIRCULATING TNF-BINDING PROTEIN IN HUMANS Mikael
Lantz,“*
Saleem Malik,2
Maurive
L. Sleviq2
Inge Olssonl
Serum samples from cancer patients receiving intravenous infusions of recombinant tumor necrosis factor (rTNF) and recombinant interferon-y (rIFN-y) were analyzed for TNF and the TNF-binding protein (TNF-BP). TNF-BP is a soluble fragment of the transmembrane TNF receptor with antagonistic effects to TNF and is released by proteolytic cleavage of the receptor. During a 60-min infusion of rTNF, peak serum levels of rTNF were observed after 30 to 60 min and a transient increase of circulating TNF-BP was observed with peak levels between 30 and 120 min. Injection of IFN-r alone did not affect the levels of TNF and TNF-BP. Thus administration of rTNF leads to release into the circulation of TNF-BP, which may modulate both systemic and local effects of TNF and influence its therapeutic efficacy.
o 1990 by W.B. Saunders Company.
Tumor necrosis factor (TNF) was initially recognized for its cytotoxic effect on tumor cells in vitro and for its capacity to cause necrosis of solid tumors in vivo. ‘v2 It is identical with a factor called cachectin,314 which suppresses the expression of anabolic enzymes in adipocytes.’ TNF/cachectin is regarded as a pleiotropic mediator of many inflammatory responses.2 It is produced by activated macrophages1,2 and one stimulus for its production is lipopolysaccharide (LPS) of gramnegative bacteria. Thus TNF/cachectin plays a major role in the pathogenesis of gram-negative endotoxic shock,6,7 which can lead to cardiovascular collapse, organ failure, and death. These effects and other harmful manifestations of TNF are reversed by passive immunization with anti-TNF monoclonal antibodies.@ Mechanisms by which the adverse effects of TNF can be suppressed may have clinical applications. A TNFbinding protein, (TNF-BP) of serum and urine has been identified.g*‘O TNF-BP was purified from urine and found to be a 30-kDa cysteine-rich glycoprotein without homologies to previously described protein sequences.’ l-l3 It prevents the binding of TNF to its cellular receptor and inhibits the biological activity of TNF:‘-13 Recent evidence suggests that TNF-BP, which binds, to TNF
‘Division of Hematology, Department of Medicine, S-22185 Lund, Sweden. ZICRF Department of Medical Oncology, St. Bartholomew’s and Homerton Hospitals, London, UK. *To whom reprint requests should be addressed at Research Department 2, E-blocket, Lund hospital, S-22185 Lund, Sweden. o 1990 by W.B. Saunders Company. 1043-4666/90/0206-0003$05.00/O KEY
402
WORDS:
ELISA/Receptor/TNF/TNF-BP
with high affinity, is a soluble fragment of the transmembrane TNF-receptor.14’15 TNF has been assessed in several predominantly phase I trials, but the overall response rates have been less than !J%.‘~-~’It has been suggested that this may be due to the relatively low doses of TNF used in clinical trials compared to the doses used in experimental models.21 We have analyzed another potential cause of the inefficacy of TNF in clinical trials, i.e., the release of a TNF inhibitor following systemic administration of TNF. Our results show that TNF infusion leads to a rapid increase of serum TNF-BP in patients with advanced cancer. RESULTS A direct ELISA to measure serum TNF was developed with a monoclonal anti-TNF antibody as the first and a polyclonal anti-TNF antibody as the second antibody. The ELISA was shown to detect both free TNF and TNF in complex with TNF-BP. To demonstrate this, TNF was diluted in TNF-BP-free serum supplemented with 1 pg/mL of purified TNF-BP. In addition, TNF was diluted in serum-from patients with chronic renal failure, who have increased levels of serum TNF-BP (100-200 ng/mL). Neither the addition of TNF-BP nor the addition of serum with high content of TNF-BP affected the standard curve for TNF (Fig. 1). We conclude that the ELISA utilized measures TNF independent of the presence of TNF-BP and thus gives a measure of total TNF present in serum samples. TNF and TNF-BP levels were determined in serum samples collected at frequent intervals after the administration of rTNF to five patients with various cancer CYTOKINE,
Vol. 2, No. 6 (November),
1990: pp 402-406
The release
B
P
of TNF-binding
protein
by TNF
/ 403
infusion of rTNF and reached a peak after 30 to 120 min (Fig. 2). In some cases the peak level of TNF-BP corresponded to the peak level of TNF, while in some cases it occurred later than for TNF. In all cases the disappearance rate was much slower for TNF-BP than for TNF. In one patient the response to infusions of rTNF was analyzed during three consecutive weeks (Fig. 3). Only rIFN-y was given the first week and it did not affect the serum levels of TNF and TNF-BP. After three weeks of treatment with rIFN-y and rTNF the peak level of serum TNF-BP in response to rTNF was increased. DISCUSSION
0 TNF hg/ml)
TNF b-@nl) Figure 1. The effect of serum TNF.
of TNF-BP
in the assay
used for determination
(A) TNF-standard diluted in TNF-BP-free serum (U) or in serum from patients with chronic renal failure containing high levels of TNF-BP (O-O). (B) TNF-standard diluted in TNF-BP-free serum (M) or in TNF-BP-free serum supplemented with 1 pg/mL TNF-BP (o--o).
diseases. These patients also received daily subcutaneous doses of rIFN-y. During a 60-min infusion of rTNF, maximal levels of serum TNF were observed after 15 to 60 min, followed by a rapid decrease to undetectable levels after approximately 2 hr (Fig. 2). The peak levels of serum TNF correlated with the dose of rTNF given. Thus patients III and IV in Fig. 2 received the lowest doses of rTNF and showed the lowest serum levels of TNF. Serum TNF-BP was found to rise during the
The major finding of this study was that administration of rTNF led to a rapid release of TNF-BP into the circulation. As TNF-BP can abrogate the cytotoxic effects of TNF in vitro,12s14the in vivo antitumor activity of rTNF may be similarly modulated. The administration of rIFN-y per se did not affect the kinetics of TNF-BP. There are now several lines of evidence that suggest that TNF-BP is a soluble fragment of a receptor for TNF. Biosynthetic labeling showed the release of TNF-BP into the supernatants in vitro. As no corresponding intracellular peptide was detected, the TNF-BP is likely to be formed and released at the cell surface.14 Molecular cloning and expression of a receptor for TNF also support the idea that the TNF-BP is derived from a transmembrane TNF receptor by proteolytic cleavage. 22-24Agents that activate protein kinase C, such as phorbol esters, downregulate the receptor25 and also induce production of TNF-BP.14 It is therefore
’1
6
60
1
0.2
20
0.1
10
0
f6
Figure 2. Results from determinations of TNF and TNF-BP in four patients (I-IV) after administration of rTNF by intravenous infusion for 1 hr with start at time point zero. All patients also received daily subcutaneous doses of IFN-7. The following doses of rTNF were received by the patients: 72 bg (I), 58.5 pg (II), 20.4 rg (III) and 20.4 pg (IV).
E 2 s
”
a4 v)
t
2
0
2
4
Time
(hours)
2
4
6
4
6
60
E 2
0
0
6
40
0.2
20
0.1
0
0
0
2 Time
(hours)
0
CYTOKINE,
404 / Lantz et al.
Vol. 2, No. 6 (November 1990: 402-406)
40 30 20 10 0 li!!cL
0
2
4
a-
8
Figure 3. Serum levels of TNF and TNF-BP as a response to weekly infusions of rTNF in one patient. Daily subcutaneous doses of rIFN-y were given throughout the four-week neriod. Infusion of rTNF was given for 1 hr with start at time point zero except for week one (upper left) when only IFN-7 was given. The following doses of rTNF were given: 61 pg week two (upper right), 92 pg week three (lower left), and 92 pg week four (lower right).
Time (hours) possible that the administration of rTNF results in activation of protein kinase C with subsequent rapid release of TNF-BP into the circulation. In this context it is of interest that we have observed that TNF itself induced production of TNF-BP in vitro.14 A similar mechanism may operate to explain the release of TNF-BP in vitro observed in the present work. Moreover, evidence exists for the development of tolerance to the toxic effect of TNF after exposure to TNF.26,27 One explanation for the induction of resistance to TNF by TNF itself could be a production of TNF-BP. Thus it is also possible that the increase of TNF-BP observed after infusion of rTNF reflects a mechanism of protection, although it remains to be shown whether TNF-BP can function as an antagonist to TNF in vivo. If so, the ratio of free versus bound rTNF may influence the therapeutic efficacy. The half-life of rTNF in the circulation is very short and the mechanisms for clearance of TNF are unknown. One explanation for the short half-life of rTNF could be that the observed increase in TNF-BP leads to formation of complexes with rTNF, which may not be detected by the assay for TNF if the antibody utilized is directed against an epitope hidden by TNF-BP. This possibility was ruled out because the assay used detected TNF independent of the presence of TNF-BP. Another explanation is that binding of rTNF to TNF-BP may lead to rapid elimination of the complexes formed. An additional possibility is that TNF is rapidly eliminated from the circulation by binding to heparin sulfate of the
extracellular matrix because TNF has an affinity for heparin.** In conclusion, administration of rTNF leads to a rapid but transient increase of TNF-BP in the circulation, which may affect the therapeutic efficacy of TNF.
MATERIALS
AND
METHODS
Materials Sepharose was from Pharmacia, Sweden. Recombinant TNF and a monoclonal antibody to TNF for ELISA were kindly provided by Dr. Gtinter Adolf, Ernst Boehringer Institute, Vienna, Austria. Recombinant human IFN-7 was kindly provided by Biogen SA (Basle, Switzerland). Recombinant TNF (specific activity 2.2 x lo6 U/mg) for the treatment of cancer patients was provided by the Asahi Chemical
Company, Japan. Native TNF-BP was purified as described.” Polyclonal antibodies to TNF and TNF-BP were made by immunizing rabbits with recombinant TNF or native TNF-
BP. The antibodies against TNF were specific and did not crossreact
with
TNF-BP
or lymphotoxin.
The antibodies
against TNF-BP did not crossreactwith TNF or lymphotoxin.
Patient Material and Collection of Serum Samples Five patients with advanced cancer were studied. The types of cancers in these patients were: malignant mesothelioma (n = 2), leiomyosarcoma (n = 2), and renal cell carci-
noma (n = 1). None of the patients were on any other drug
The release of TNF-binding
or had signs of infective complications during the study period. The patients studied were entered into a phase I trial of rTNF and rIFN-7. The basic design of the study was to treat patients with rIFN-y only for the first week (100 fig/m*, daily subcutaneously for five days). In subsequent weeks, patients received rTNF as a 60-min intravenous infusion three
treatment,
times a week in an escalating dose-schedule, in addition to daily subcutaneous rIFN-y. Serum samples were collected
before, at the beginning, and at frequent intervals after the infusion of rTNF.
Preparation
of Serum Free of TNF and TNF-BP
Antibodies to TNF-BP were coupled to Sepharose by the cyanogen
bromide
technique.*’
Ten
milliliters
of swollen
anti-TNF-BP-Sepharose were applied to a column and equilibrated with column buffer (0.15 mol/L NaCl, 5 mmol/L Hepes, pH 7.4). Pooled serum from healthy individuals was applied to the column at a flow rate of 10 mL/hr. Fractions of 10 mL were collected and assayedwith the ELISA technique
to make sure that they did not contain TNF or TNF-BP.
Assay for Determination
of Serum TNF-BP
A competitive inhibitory ELISA, using a polyclonal antibody, was developed for the measurement of TNF-BP in serum. Samples of 50 ~1were loaded in 96-well round-bottom incubation plates (Nunc 2-62170). A 50 ~1 aliquot of antiTNF-BP (polyclonal antibody) diluted 1:4000 in incubation buffer (0.1 mol/L NaCl, 0.05 mol/L NaH,PO,, 0.05% Tween20, adjusted to pH 7.5 with NaOH) was added to each well
followed by incubation over night at 4‘C. In parallel, 96-well flat-bottom immunoplates (Nunc 4-394459) were coated with 100 ~1 per well of TNF-BP NaHCO,, and incubated
diluted to 7.5 ng/mL in 0.1 mol/L overnight at room temperature
bovine serum albumin
was added to each well.
The goat anti-rabbit antibody was conjugated to biotin (Vectastain ABC kit, AK 5001, Vector Laboratories). The plates were incubated for 1 hr at room temperature followed by three washes. To enhance the signal, 100 ~1 of a solution containing
avidin-biotin complexes with alkaline phosphatase conjugated to biotin (Vectastain ABC kit, AK 5001, Vector Laboratories) was added to each well followed by incubation of the plates for 1 hr at room temperature and washed three times. 100 ~1 of phosphatase
substrate
(Sigma
104) dissolved
in 1 mol/L
diethanolamine buffer, pH 9.8, supplemented with 0.5 pmol/L MgCl,, was added to each well. The absorbance was measured at 405 nm in an automatic
Titertek
multiscan
ELISA
plate
reader. Values were calculated from a standard curve based on freshly
prepared
dilutions
TNF-BP-free serum.
of TNF-BP
(0.3-30
ng/mL)
of Serum TNF
A direct ELISA was utilized for the measurement of serum TNF. Nunc 96-well immunoplates (4-39445) were coated with a monoclonal antibody to TNF diluted to 2.5 pg/mL (40 pi/well) in 0.1 mol/L NaHCO, and incubated for at least 3 hr at room temperature, followed by three washesin washing solution (0.15 mol/L NaCl, 0.05% Tween-20). Samples of 100 ~1were loaded in triplicate and incubated overnight at 4OC followed by three washes in washing solution. An aliquot of 100 ~1 of a polyclonal antibody to TNF, diluted 1:3,600 in incubation buffer (0.1 mol/L NaCl, 0.05 mol/L NaH,PO,, 0.05% Tween-20, adjusted to pH 7.5 with NaOH) supplemented with 1 mg/mL bovine serum albumin, was added to each well. Plates were incubated for 3 hr at room temperature followed by three washes in washing solution. A peroxidase-conjugated
goat anti-rabbit
antibody,
(170-6513,
BioRad, Richmond, CA, USA) diluted 1:2,500 in incubation buffer and supplemented with 1 mg/mL of bovine serum albumin was prepared, and an aliquot of 100 ~1was added to each well. Plates were incubated for 1 hr at room temperature, washed three times in washing solution followed by addition of 100 ~1 of tetramethyl benzidine peroxidase (172-1066, BioRad, Richmond, CA, USA) to each well. The absorbance was measured at 660 nm in an automatic Titertek ELISA plate reader. Values were calculated from a standard curve basedon freshly prepared dilutions of TNF (0.02 to 20 ng/mL) in TNF-free serum. Acknowledgments This work was supported by the Swedish Cancer Society, John and Augusta Persson Foundation, Alfred dsterlund Foundation, Kamprad Foundation, and the Medical Faculty of Lund. REFERENCES
followed by three washes in washing solution (0.15 mol/L NaCl, 0.05% Tween-20). The content of the incubation plate was transferred to the plate coated with TNF-BP, which was incubated for three hr at 4°C and washed three times. An aliquot of 100 ~1 of goat anti-rabbit antibody (secondary antibody), diluted 1:200 in incubation buffer supplemented with 1 mg/mL
Assay for Determination
protein by TNF / 405
in
1. Carswell EA, Old LJ, Kassel RL, Green S, Fiore N, Williamsson B (1975) An endotoxin-induced serum factor that causes necrosis of tumors. Proc Nat1 Acad Sci USA 72:3666-3670. 2. Old LJ (1987) Tumor necrosis factor. Polypeptide mediator network. Nature 326:330-331. 3. Beutler B, Greenwald D, Hulmes JD, Chang M, Pan YCE, Mathison J, Ulevitch R, Cerami A (1985) Identity of tumor necrosis factor and the macrophage secreted factor cachectin. Nature 3 16:552554. 4. Sherry B, Cerami A (1988) Cachectin/tumor necrosis factor exerts endocrine, paracrine, and autocrine control of inflammatory responses. J Cell Biol 107:1269-1277. 5. Torti FM, Dieckmann B, Beutler B, Cerami A, Ringold GM (1985) A macrophage factor inhibits adipocyte gene expression: an in vitro model of cachexia. Science 229:867-869. 6. Beutler B, Milsark IW, Cerami A (1985) Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229:869-87 1. 7. Tracey KJ, Fong Y, Hesse DG, Manogue KR, Lee AT, Kuo GC, Lowry SF, Cerami A (1987) Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteraemia. Nature 330:662-664. 8. Piguet PF, Grau GE, Allet B, Vassalli P (1987) Tumor necrosis factorlcachectin is an effector of skin and gut lesions of the acute phase of graft-vs.-host disease. J Exp Med 166:1280-1289.
406 / Lantz et al. 9. Peetre C, Thysell H, Grubb A, Olsson I (1988) A tumor necrosis factor binding protein is present in human biological fluids. Eur J Haematol41:414-419. 10. Seckinger P, Isaaz S, Dayer JM (1988) A human inhibitor for tumor necrosis factor. J Exp Med 167:1511-1516. 11. Olsson I, Lantz M, Nilsson E, Peetre C, Thysell H, Grubb A, Adolf G (1989) Isolation and characterization of a tumor necrosis factor binding protein from urine. Eur J Haematol42:270-275. 12. Engelmann H, Aderka D, Rubinstein M, Rotman D, Wallach D (1989) A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem 264: 11974- 11980. 13. Seckinger P, Isaaz S, Dayer JM (1989) Purification and biologic characterization of a specific tumor necrosis factor inhibitor. J Biol Chem 264:11966-11973. 14. Lantz M, Gullberg U, Nilsson E, Olsson I (in press) Characterization in vitro of a human tumor necrosis factor-binding protein: A soluble form of a tumor necrosis factor receptor. J Clin Invest 15. Engelmann H, Novick D, Wallach D (1989) Two tumor necrosis factor-binding proteins purified from human urine. Evidence for immunological cross-reactivity with cell surface tumor necrosis factor receptors. J Biol Chem 265:1531-1536. 16. Feinberg B, Kurzrock R, Talpaz M, Blick M, Saks M, Gutterman JU (1988) A phase I trial of intravenously administrated recombinant tumor necrosis factor-alpha in cancer patients. J Clin Oncol6:1328-1334. 17. Blick M, Sherwin SA, Rosenblum M, Gutterman (1987) Phase I study of recombinant tumor necrosis factor in cancer patients. Cancer Res 47:2986-2989. 18. Creaven PJ, Plager JE, Dupere S, Huben RP, Takita H, Mittelman A, Proefrock A (1987) Phase I clinical trial of recombinant human tumor necrosis factor. Cancer Chemother Pharmacol 20:223229. 19. Kimura K, Taguchi T, Urushizaki I, Ohno R, Abe 0, Furue H, Hattori T, Ichihashi H, Inoguchi K, Majima H, Niitani H, Ota K, Saito T, Suga S (1987) Phase I study of recombinant human tumor necrosis factor. Cancer Chemother Pharmacol20:230-235. 20. Spriggs DR, Sherman ML, Michie H, Arthur KAA, Imam-
CYTOKINE,
Vol. 2, No. 6 (November 1990: 402-406)
ura K, Wilmore D, Frei E, Kufe DW (1988) Recombinant human tumor necrosis factor administrated as a 24-hour intravenous infusion. A phase I and pharmacologic study. J Nat1 Cancer Inst 80:1039-1044. 21. Frei E, Spriggs D (1989) Tumor necrosis factor: still a promising agent. J Clin Oncol7:291-294. 22. Himmler A, Maurer-Fogy I, Kr6nke M, Scheurich P, Pfizenmaier K, Lantz M, Olsson I, Hauptmann R, Stratowa C, Adolf GR (1990) Molecular cloning and expression of human and rat tumor necrosis factor receptor chain (~60) and its soluble derivative, tumor necrosis factor-binding protein. DNA and Cell Biology 9:705-7 15. 23. Loetscher H, Pan YCE, Lahm HW, Gentz R, Brockhaus M, Tabuchi H, Lesslaer W (1990) Molecular cloning and expression of the 55 kd tumor necrosis factor receptor. Cell 61:352-359. 24. Schall TJ, Lewis M, Koller KJ, Lee A, Rice GC, Wong GHW, Gatanaga T, Granger GA, Lentz R, Raab H, Kohr WJ, Goeddel DV (1990) Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell 61:361-370. 25. Gullberg U, Lantz M, Nilsson E, Peetre C, Adolf G, Olsson I (1987) Characterization of a relationship between the T-lymphocyte derived differentiation inducing factor (DIF) and lymphotoxin: A common receptor system for DIF, lymphotoxin and tumor necrosis factor downregulated by phorbol diesters. Eur J Haematol 39:241251. 26. Wallach D, Holtmann H, Engelmann H, Nophar Y (1988) Sensitization and desensitization to lethal effects of tumor necrosis factor and IL-l. J Immunol 140:2994-2999. 27. Galanos C, Freudenberg MA, Coumbos A, Matsuura M, Lehmann V, Bartholeyns J (1988) Induction of lethality and tolerance by endotoxin are mediated by macrophages through tumor necrosis factor. In Bonavida B, Gifford GE, Kirchner H, Old LJ (eds) Tumor Necrosis Factor/Cachectin and Related Cytokines, Karger, Basel, pp 114-127. 28. Lantz M, Thysell H, Nilsson E, Olsson I (submitted) On the binding of tumor necrosis factor (TNF) to heparin and the release in vivo of the TNF-binding protein by heparin. 29. Axen R, Porath J, Ernback S (1967) Chemical coupling of peptides and proteins to polysaccharides by means of cyanogen halides. Nature 214:1302-1304.
Lantz,“*
Saleem Malik,2
Maurive
L. Sleviq2
Inge Olssonl
Serum samples from cancer patients receiving intravenous infusions of recombinant tumor necrosis factor (rTNF) and recombinant interferon-y (rIFN-y) were analyzed for TNF and the TNF-binding protein (TNF-BP). TNF-BP is a soluble fragment of the transmembrane TNF receptor with antagonistic effects to TNF and is released by proteolytic cleavage of the receptor. During a 60-min infusion of rTNF, peak serum levels of rTNF were observed after 30 to 60 min and a transient increase of circulating TNF-BP was observed with peak levels between 30 and 120 min. Injection of IFN-r alone did not affect the levels of TNF and TNF-BP. Thus administration of rTNF leads to release into the circulation of TNF-BP, which may modulate both systemic and local effects of TNF and influence its therapeutic efficacy.
o 1990 by W.B. Saunders Company.
Tumor necrosis factor (TNF) was initially recognized for its cytotoxic effect on tumor cells in vitro and for its capacity to cause necrosis of solid tumors in vivo. ‘v2 It is identical with a factor called cachectin,314 which suppresses the expression of anabolic enzymes in adipocytes.’ TNF/cachectin is regarded as a pleiotropic mediator of many inflammatory responses.2 It is produced by activated macrophages1,2 and one stimulus for its production is lipopolysaccharide (LPS) of gramnegative bacteria. Thus TNF/cachectin plays a major role in the pathogenesis of gram-negative endotoxic shock,6,7 which can lead to cardiovascular collapse, organ failure, and death. These effects and other harmful manifestations of TNF are reversed by passive immunization with anti-TNF monoclonal antibodies.@ Mechanisms by which the adverse effects of TNF can be suppressed may have clinical applications. A TNFbinding protein, (TNF-BP) of serum and urine has been identified.g*‘O TNF-BP was purified from urine and found to be a 30-kDa cysteine-rich glycoprotein without homologies to previously described protein sequences.’ l-l3 It prevents the binding of TNF to its cellular receptor and inhibits the biological activity of TNF:‘-13 Recent evidence suggests that TNF-BP, which binds, to TNF
‘Division of Hematology, Department of Medicine, S-22185 Lund, Sweden. ZICRF Department of Medical Oncology, St. Bartholomew’s and Homerton Hospitals, London, UK. *To whom reprint requests should be addressed at Research Department 2, E-blocket, Lund hospital, S-22185 Lund, Sweden. o 1990 by W.B. Saunders Company. 1043-4666/90/0206-0003$05.00/O KEY
402
WORDS:
ELISA/Receptor/TNF/TNF-BP
with high affinity, is a soluble fragment of the transmembrane TNF-receptor.14’15 TNF has been assessed in several predominantly phase I trials, but the overall response rates have been less than !J%.‘~-~’It has been suggested that this may be due to the relatively low doses of TNF used in clinical trials compared to the doses used in experimental models.21 We have analyzed another potential cause of the inefficacy of TNF in clinical trials, i.e., the release of a TNF inhibitor following systemic administration of TNF. Our results show that TNF infusion leads to a rapid increase of serum TNF-BP in patients with advanced cancer. RESULTS A direct ELISA to measure serum TNF was developed with a monoclonal anti-TNF antibody as the first and a polyclonal anti-TNF antibody as the second antibody. The ELISA was shown to detect both free TNF and TNF in complex with TNF-BP. To demonstrate this, TNF was diluted in TNF-BP-free serum supplemented with 1 pg/mL of purified TNF-BP. In addition, TNF was diluted in serum-from patients with chronic renal failure, who have increased levels of serum TNF-BP (100-200 ng/mL). Neither the addition of TNF-BP nor the addition of serum with high content of TNF-BP affected the standard curve for TNF (Fig. 1). We conclude that the ELISA utilized measures TNF independent of the presence of TNF-BP and thus gives a measure of total TNF present in serum samples. TNF and TNF-BP levels were determined in serum samples collected at frequent intervals after the administration of rTNF to five patients with various cancer CYTOKINE,
Vol. 2, No. 6 (November),
1990: pp 402-406
The release
B
P
of TNF-binding
protein
by TNF
/ 403
infusion of rTNF and reached a peak after 30 to 120 min (Fig. 2). In some cases the peak level of TNF-BP corresponded to the peak level of TNF, while in some cases it occurred later than for TNF. In all cases the disappearance rate was much slower for TNF-BP than for TNF. In one patient the response to infusions of rTNF was analyzed during three consecutive weeks (Fig. 3). Only rIFN-y was given the first week and it did not affect the serum levels of TNF and TNF-BP. After three weeks of treatment with rIFN-y and rTNF the peak level of serum TNF-BP in response to rTNF was increased. DISCUSSION
0 TNF hg/ml)
TNF b-@nl) Figure 1. The effect of serum TNF.
of TNF-BP
in the assay
used for determination
(A) TNF-standard diluted in TNF-BP-free serum (U) or in serum from patients with chronic renal failure containing high levels of TNF-BP (O-O). (B) TNF-standard diluted in TNF-BP-free serum (M) or in TNF-BP-free serum supplemented with 1 pg/mL TNF-BP (o--o).
diseases. These patients also received daily subcutaneous doses of rIFN-y. During a 60-min infusion of rTNF, maximal levels of serum TNF were observed after 15 to 60 min, followed by a rapid decrease to undetectable levels after approximately 2 hr (Fig. 2). The peak levels of serum TNF correlated with the dose of rTNF given. Thus patients III and IV in Fig. 2 received the lowest doses of rTNF and showed the lowest serum levels of TNF. Serum TNF-BP was found to rise during the
The major finding of this study was that administration of rTNF led to a rapid release of TNF-BP into the circulation. As TNF-BP can abrogate the cytotoxic effects of TNF in vitro,12s14the in vivo antitumor activity of rTNF may be similarly modulated. The administration of rIFN-y per se did not affect the kinetics of TNF-BP. There are now several lines of evidence that suggest that TNF-BP is a soluble fragment of a receptor for TNF. Biosynthetic labeling showed the release of TNF-BP into the supernatants in vitro. As no corresponding intracellular peptide was detected, the TNF-BP is likely to be formed and released at the cell surface.14 Molecular cloning and expression of a receptor for TNF also support the idea that the TNF-BP is derived from a transmembrane TNF receptor by proteolytic cleavage. 22-24Agents that activate protein kinase C, such as phorbol esters, downregulate the receptor25 and also induce production of TNF-BP.14 It is therefore
’1
6
60
1
0.2
20
0.1
10
0
f6
Figure 2. Results from determinations of TNF and TNF-BP in four patients (I-IV) after administration of rTNF by intravenous infusion for 1 hr with start at time point zero. All patients also received daily subcutaneous doses of IFN-7. The following doses of rTNF were received by the patients: 72 bg (I), 58.5 pg (II), 20.4 rg (III) and 20.4 pg (IV).
E 2 s
”
a4 v)
t
2
0
2
4
Time
(hours)
2
4
6
4
6
60
E 2
0
0
6
40
0.2
20
0.1
0
0
0
2 Time
(hours)
0
CYTOKINE,
404 / Lantz et al.
Vol. 2, No. 6 (November 1990: 402-406)
40 30 20 10 0 li!!cL
0
2
4
a-
8
Figure 3. Serum levels of TNF and TNF-BP as a response to weekly infusions of rTNF in one patient. Daily subcutaneous doses of rIFN-y were given throughout the four-week neriod. Infusion of rTNF was given for 1 hr with start at time point zero except for week one (upper left) when only IFN-7 was given. The following doses of rTNF were given: 61 pg week two (upper right), 92 pg week three (lower left), and 92 pg week four (lower right).
Time (hours) possible that the administration of rTNF results in activation of protein kinase C with subsequent rapid release of TNF-BP into the circulation. In this context it is of interest that we have observed that TNF itself induced production of TNF-BP in vitro.14 A similar mechanism may operate to explain the release of TNF-BP in vitro observed in the present work. Moreover, evidence exists for the development of tolerance to the toxic effect of TNF after exposure to TNF.26,27 One explanation for the induction of resistance to TNF by TNF itself could be a production of TNF-BP. Thus it is also possible that the increase of TNF-BP observed after infusion of rTNF reflects a mechanism of protection, although it remains to be shown whether TNF-BP can function as an antagonist to TNF in vivo. If so, the ratio of free versus bound rTNF may influence the therapeutic efficacy. The half-life of rTNF in the circulation is very short and the mechanisms for clearance of TNF are unknown. One explanation for the short half-life of rTNF could be that the observed increase in TNF-BP leads to formation of complexes with rTNF, which may not be detected by the assay for TNF if the antibody utilized is directed against an epitope hidden by TNF-BP. This possibility was ruled out because the assay used detected TNF independent of the presence of TNF-BP. Another explanation is that binding of rTNF to TNF-BP may lead to rapid elimination of the complexes formed. An additional possibility is that TNF is rapidly eliminated from the circulation by binding to heparin sulfate of the
extracellular matrix because TNF has an affinity for heparin.** In conclusion, administration of rTNF leads to a rapid but transient increase of TNF-BP in the circulation, which may affect the therapeutic efficacy of TNF.
MATERIALS
AND
METHODS
Materials Sepharose was from Pharmacia, Sweden. Recombinant TNF and a monoclonal antibody to TNF for ELISA were kindly provided by Dr. Gtinter Adolf, Ernst Boehringer Institute, Vienna, Austria. Recombinant human IFN-7 was kindly provided by Biogen SA (Basle, Switzerland). Recombinant TNF (specific activity 2.2 x lo6 U/mg) for the treatment of cancer patients was provided by the Asahi Chemical
Company, Japan. Native TNF-BP was purified as described.” Polyclonal antibodies to TNF and TNF-BP were made by immunizing rabbits with recombinant TNF or native TNF-
BP. The antibodies against TNF were specific and did not crossreact
with
TNF-BP
or lymphotoxin.
The antibodies
against TNF-BP did not crossreactwith TNF or lymphotoxin.
Patient Material and Collection of Serum Samples Five patients with advanced cancer were studied. The types of cancers in these patients were: malignant mesothelioma (n = 2), leiomyosarcoma (n = 2), and renal cell carci-
noma (n = 1). None of the patients were on any other drug
The release of TNF-binding
or had signs of infective complications during the study period. The patients studied were entered into a phase I trial of rTNF and rIFN-7. The basic design of the study was to treat patients with rIFN-y only for the first week (100 fig/m*, daily subcutaneously for five days). In subsequent weeks, patients received rTNF as a 60-min intravenous infusion three
treatment,
times a week in an escalating dose-schedule, in addition to daily subcutaneous rIFN-y. Serum samples were collected
before, at the beginning, and at frequent intervals after the infusion of rTNF.
Preparation
of Serum Free of TNF and TNF-BP
Antibodies to TNF-BP were coupled to Sepharose by the cyanogen
bromide
technique.*’
Ten
milliliters
of swollen
anti-TNF-BP-Sepharose were applied to a column and equilibrated with column buffer (0.15 mol/L NaCl, 5 mmol/L Hepes, pH 7.4). Pooled serum from healthy individuals was applied to the column at a flow rate of 10 mL/hr. Fractions of 10 mL were collected and assayedwith the ELISA technique
to make sure that they did not contain TNF or TNF-BP.
Assay for Determination
of Serum TNF-BP
A competitive inhibitory ELISA, using a polyclonal antibody, was developed for the measurement of TNF-BP in serum. Samples of 50 ~1were loaded in 96-well round-bottom incubation plates (Nunc 2-62170). A 50 ~1 aliquot of antiTNF-BP (polyclonal antibody) diluted 1:4000 in incubation buffer (0.1 mol/L NaCl, 0.05 mol/L NaH,PO,, 0.05% Tween20, adjusted to pH 7.5 with NaOH) was added to each well
followed by incubation over night at 4‘C. In parallel, 96-well flat-bottom immunoplates (Nunc 4-394459) were coated with 100 ~1 per well of TNF-BP NaHCO,, and incubated
diluted to 7.5 ng/mL in 0.1 mol/L overnight at room temperature
bovine serum albumin
was added to each well.
The goat anti-rabbit antibody was conjugated to biotin (Vectastain ABC kit, AK 5001, Vector Laboratories). The plates were incubated for 1 hr at room temperature followed by three washes. To enhance the signal, 100 ~1 of a solution containing
avidin-biotin complexes with alkaline phosphatase conjugated to biotin (Vectastain ABC kit, AK 5001, Vector Laboratories) was added to each well followed by incubation of the plates for 1 hr at room temperature and washed three times. 100 ~1 of phosphatase
substrate
(Sigma
104) dissolved
in 1 mol/L
diethanolamine buffer, pH 9.8, supplemented with 0.5 pmol/L MgCl,, was added to each well. The absorbance was measured at 405 nm in an automatic
Titertek
multiscan
ELISA
plate
reader. Values were calculated from a standard curve based on freshly
prepared
dilutions
TNF-BP-free serum.
of TNF-BP
(0.3-30
ng/mL)
of Serum TNF
A direct ELISA was utilized for the measurement of serum TNF. Nunc 96-well immunoplates (4-39445) were coated with a monoclonal antibody to TNF diluted to 2.5 pg/mL (40 pi/well) in 0.1 mol/L NaHCO, and incubated for at least 3 hr at room temperature, followed by three washesin washing solution (0.15 mol/L NaCl, 0.05% Tween-20). Samples of 100 ~1were loaded in triplicate and incubated overnight at 4OC followed by three washes in washing solution. An aliquot of 100 ~1 of a polyclonal antibody to TNF, diluted 1:3,600 in incubation buffer (0.1 mol/L NaCl, 0.05 mol/L NaH,PO,, 0.05% Tween-20, adjusted to pH 7.5 with NaOH) supplemented with 1 mg/mL bovine serum albumin, was added to each well. Plates were incubated for 3 hr at room temperature followed by three washes in washing solution. A peroxidase-conjugated
goat anti-rabbit
antibody,
(170-6513,
BioRad, Richmond, CA, USA) diluted 1:2,500 in incubation buffer and supplemented with 1 mg/mL of bovine serum albumin was prepared, and an aliquot of 100 ~1was added to each well. Plates were incubated for 1 hr at room temperature, washed three times in washing solution followed by addition of 100 ~1 of tetramethyl benzidine peroxidase (172-1066, BioRad, Richmond, CA, USA) to each well. The absorbance was measured at 660 nm in an automatic Titertek ELISA plate reader. Values were calculated from a standard curve basedon freshly prepared dilutions of TNF (0.02 to 20 ng/mL) in TNF-free serum. Acknowledgments This work was supported by the Swedish Cancer Society, John and Augusta Persson Foundation, Alfred dsterlund Foundation, Kamprad Foundation, and the Medical Faculty of Lund. REFERENCES
followed by three washes in washing solution (0.15 mol/L NaCl, 0.05% Tween-20). The content of the incubation plate was transferred to the plate coated with TNF-BP, which was incubated for three hr at 4°C and washed three times. An aliquot of 100 ~1 of goat anti-rabbit antibody (secondary antibody), diluted 1:200 in incubation buffer supplemented with 1 mg/mL
Assay for Determination
protein by TNF / 405
in
1. Carswell EA, Old LJ, Kassel RL, Green S, Fiore N, Williamsson B (1975) An endotoxin-induced serum factor that causes necrosis of tumors. Proc Nat1 Acad Sci USA 72:3666-3670. 2. Old LJ (1987) Tumor necrosis factor. Polypeptide mediator network. Nature 326:330-331. 3. Beutler B, Greenwald D, Hulmes JD, Chang M, Pan YCE, Mathison J, Ulevitch R, Cerami A (1985) Identity of tumor necrosis factor and the macrophage secreted factor cachectin. Nature 3 16:552554. 4. Sherry B, Cerami A (1988) Cachectin/tumor necrosis factor exerts endocrine, paracrine, and autocrine control of inflammatory responses. J Cell Biol 107:1269-1277. 5. Torti FM, Dieckmann B, Beutler B, Cerami A, Ringold GM (1985) A macrophage factor inhibits adipocyte gene expression: an in vitro model of cachexia. Science 229:867-869. 6. Beutler B, Milsark IW, Cerami A (1985) Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229:869-87 1. 7. Tracey KJ, Fong Y, Hesse DG, Manogue KR, Lee AT, Kuo GC, Lowry SF, Cerami A (1987) Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteraemia. Nature 330:662-664. 8. Piguet PF, Grau GE, Allet B, Vassalli P (1987) Tumor necrosis factorlcachectin is an effector of skin and gut lesions of the acute phase of graft-vs.-host disease. J Exp Med 166:1280-1289.
406 / Lantz et al. 9. Peetre C, Thysell H, Grubb A, Olsson I (1988) A tumor necrosis factor binding protein is present in human biological fluids. Eur J Haematol41:414-419. 10. Seckinger P, Isaaz S, Dayer JM (1988) A human inhibitor for tumor necrosis factor. J Exp Med 167:1511-1516. 11. Olsson I, Lantz M, Nilsson E, Peetre C, Thysell H, Grubb A, Adolf G (1989) Isolation and characterization of a tumor necrosis factor binding protein from urine. Eur J Haematol42:270-275. 12. Engelmann H, Aderka D, Rubinstein M, Rotman D, Wallach D (1989) A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem 264: 11974- 11980. 13. Seckinger P, Isaaz S, Dayer JM (1989) Purification and biologic characterization of a specific tumor necrosis factor inhibitor. J Biol Chem 264:11966-11973. 14. Lantz M, Gullberg U, Nilsson E, Olsson I (in press) Characterization in vitro of a human tumor necrosis factor-binding protein: A soluble form of a tumor necrosis factor receptor. J Clin Invest 15. Engelmann H, Novick D, Wallach D (1989) Two tumor necrosis factor-binding proteins purified from human urine. Evidence for immunological cross-reactivity with cell surface tumor necrosis factor receptors. J Biol Chem 265:1531-1536. 16. Feinberg B, Kurzrock R, Talpaz M, Blick M, Saks M, Gutterman JU (1988) A phase I trial of intravenously administrated recombinant tumor necrosis factor-alpha in cancer patients. J Clin Oncol6:1328-1334. 17. Blick M, Sherwin SA, Rosenblum M, Gutterman (1987) Phase I study of recombinant tumor necrosis factor in cancer patients. Cancer Res 47:2986-2989. 18. Creaven PJ, Plager JE, Dupere S, Huben RP, Takita H, Mittelman A, Proefrock A (1987) Phase I clinical trial of recombinant human tumor necrosis factor. Cancer Chemother Pharmacol 20:223229. 19. Kimura K, Taguchi T, Urushizaki I, Ohno R, Abe 0, Furue H, Hattori T, Ichihashi H, Inoguchi K, Majima H, Niitani H, Ota K, Saito T, Suga S (1987) Phase I study of recombinant human tumor necrosis factor. Cancer Chemother Pharmacol20:230-235. 20. Spriggs DR, Sherman ML, Michie H, Arthur KAA, Imam-
CYTOKINE,
Vol. 2, No. 6 (November 1990: 402-406)
ura K, Wilmore D, Frei E, Kufe DW (1988) Recombinant human tumor necrosis factor administrated as a 24-hour intravenous infusion. A phase I and pharmacologic study. J Nat1 Cancer Inst 80:1039-1044. 21. Frei E, Spriggs D (1989) Tumor necrosis factor: still a promising agent. J Clin Oncol7:291-294. 22. Himmler A, Maurer-Fogy I, Kr6nke M, Scheurich P, Pfizenmaier K, Lantz M, Olsson I, Hauptmann R, Stratowa C, Adolf GR (1990) Molecular cloning and expression of human and rat tumor necrosis factor receptor chain (~60) and its soluble derivative, tumor necrosis factor-binding protein. DNA and Cell Biology 9:705-7 15. 23. Loetscher H, Pan YCE, Lahm HW, Gentz R, Brockhaus M, Tabuchi H, Lesslaer W (1990) Molecular cloning and expression of the 55 kd tumor necrosis factor receptor. Cell 61:352-359. 24. Schall TJ, Lewis M, Koller KJ, Lee A, Rice GC, Wong GHW, Gatanaga T, Granger GA, Lentz R, Raab H, Kohr WJ, Goeddel DV (1990) Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell 61:361-370. 25. Gullberg U, Lantz M, Nilsson E, Peetre C, Adolf G, Olsson I (1987) Characterization of a relationship between the T-lymphocyte derived differentiation inducing factor (DIF) and lymphotoxin: A common receptor system for DIF, lymphotoxin and tumor necrosis factor downregulated by phorbol diesters. Eur J Haematol 39:241251. 26. Wallach D, Holtmann H, Engelmann H, Nophar Y (1988) Sensitization and desensitization to lethal effects of tumor necrosis factor and IL-l. J Immunol 140:2994-2999. 27. Galanos C, Freudenberg MA, Coumbos A, Matsuura M, Lehmann V, Bartholeyns J (1988) Induction of lethality and tolerance by endotoxin are mediated by macrophages through tumor necrosis factor. In Bonavida B, Gifford GE, Kirchner H, Old LJ (eds) Tumor Necrosis Factor/Cachectin and Related Cytokines, Karger, Basel, pp 114-127. 28. Lantz M, Thysell H, Nilsson E, Olsson I (submitted) On the binding of tumor necrosis factor (TNF) to heparin and the release in vivo of the TNF-binding protein by heparin. 29. Axen R, Porath J, Ernback S (1967) Chemical coupling of peptides and proteins to polysaccharides by means of cyanogen halides. Nature 214:1302-1304.
