A flow cytometric and immunofluorescence microscopic study of tumor necrosis factor production and...

CELLULAR IMMUNOLOGY 122,405-415 (1989)

A Flow Cytometric and lmmunofluorescence Microscopic Study of Tumor Necrosis Factor Production and Localization

in Human Monocytes

EVA HOFSLI,’ ODDMUND BARRE,~ UNNI NONSTAD, AND TERJE ESPEVIK

Institute of Cancer Research, University of Trondheim, N-7006 Trondheim. Norway

Received September 28,19X3; accepted May 7, I989

The production and localization of tumor necrosis factor (TNF) in human monocytes were investigated by using monoclonal and polyclonal antibodies against recombinant human TNF together with flow cytometry and immunofluorescence microscopy. Lipopolysaccharide (LPS) induced a rapid and transient accumulation of TNF in perinuclear vesicles which was detected 20 min after the addition of LPS. The fluorescence intensity of the vesicles peaked at 40 min of LPS exposure, concomitantly with the release of TNF into the medium. Thus, our results indicate that the secretion of TNF is typical for secretory proteins as it involves passage through the secretory apparatus. Additional studies demonstrated that plasma membrane-associated TNF could not be detected in live monocytes not exposed to LPS. However, alter 90 min with LPS, a small population of monocytes expressed membrane-associated TNF, and by 24 hr approximately 50% of the monocytes displayed TNF on the plasma membrane. Furthermore, our results indicate that plasma membrane-associated TNF does not represent released TNF bound back to its own receptor. Thus, our findings support the view that TNF exists as a surface transmembrane protein in LPS-stimulated monocytes. 0 1989 Academic PESS, IX.

INTRODUCTION

Tumor necrosis factor (TNF) (also called TNF-cz or cachectin) was initially described as an antitumor activity present in the serum of endotoxin-treated mice (1). It is now known that, in addition to its direct cytotoxic/cytostatic effect against tumor cells, TNF can express multiple immunomodulatory activities (2-6). Monocytes produce large quantities of TNF in response to lipopolysaccharide (LPS), which seems to stimulate both translation and transcription of TNF-mRNA (7). Previous studies have shown that TNF activity is associated with the plasma membrane of activated monocytes (8), and that TNF may exist as a 26-kDa cell surface transmembrane

’ To whom correspondence should be addressed. ’ Present address: European Molecular Biology Laboratory, Postfact 10.2209, D-6900 Heidelberg, Ger-

many. 3 Abbreviations used: TNF, tumor necrosis factor; rTNF, recombinant human tumor necrosis factor;

rLT, recombinant human lymphotoxin; LPS, lipopolysaccharide; nRIgG, normal rabbit IgG; FITC, fluo- rescein isothiocyanate.

405

0008~8749189 $3.00 Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved.

406 HOFSLI ET AL.

protein (9). However, the mechanism of TNF secretion is still unclear as it recently has been reported not to be typical for that of secretory proteins (9).

The purpose of the present study was to examine the kinetics and localization of TNF production in LPS-stimulated human monocytes using highly specific antibodies against recombinant human TNF (rTNF) together with flow cytometry and immunofluorescence microscopy. The results demonstrate that LPS induces a rapid and transient accumulation of TNF in distinct perinuclear vesicles which is detected before TNF is released into the medium. Furthermore, plasma membrane-associated TNF was detected only after the peak of intracellular TNF had occurred, indicating that TNF may be expressed as a surface transmembrane protein following the degranulation of secretory vesicles.

MATERIALS AND METHODS

Cytokines and antibodies. Recombinant human TNF and recombinant human lymphotoxin (rLT, also called TNF-@) (Genentech, South San Francisco, CA) were generously provided by Dr. G. R. Adolf, Boehringer Ingelheim, Vienna, Austria. Affinity-purified polyclonal anti-TNF antibodies were prepared by passing ammo- nium sulphate-precipitated antiserum from a rabbit immunized with rTNF over a column of cyanogen bromide-activated sepharose 4B (Pharmacia, Uppsala, Sweden) conjugated with rTNF (8). Bound anti-TNF antibodies were eluted with 0.2 M gly- tine-HCl, 0.5 MNaCl, pH 2.8, and dialyzed against phosphate-buffered saline (PBS), pH 7.4, before use. The anti-TNF antibodies neutralized the cytotoxic activity of rTNF, but not of rLT, in the WEHI 164 clone 13 bioassay ( 10). Supernatants from a hybridoma (6Hll) secreting monoclonal antibodies against TNF were also used (T. Espevik, unpublished results). This hybridoma was made by fusing spleen cells from Balb/C mice immunized with rTNF with NSO plasmacytoma cells generously provided by Dr. Z. Eshhar, Weizmann Institute of Science, Israel. The 6H 11 hybridoma secretes IgGi antibodies which completely neutralize the cytotoxic activity of rTNF in the WEHI 164 clone 13 bioassay (10). The monoclonal antibody htr-9b (IgGJ against the TNF receptor (11, 12) was obtained from M. Brockhaus, Hoffmann- LaRoche, Basel, Switzerland. Normal rabbit IgG (nRIgG) was obtained from Cappel (Malvern, PA). FITC-conjugated affinity-purified F(ab’)2 of goat anti-rabbit IgG (H + L) was purchased from Zymed, Inc. (South San Francisco, CA). FITC-conjugated goat anti-mouse IgG was purchased from Becton-Dickinson (Mountain View, CA).

Monocyte isolation and culture. Human peripheral blood mononuclear cells (PBMC) were isolated from normal huffy coats (Blood Bank, University of Trond- heim, Norway) by Lymphoprep (Nycomed, Oslo, Norway) sedimentation ( 13). The purified PBMC were suspended in complete medium consisting of RPM1 1640 (GIBCO, Paisley, UK) containing 25% pooled human A+ serum (Blood Bank, Uni- versity of Trondheim), 2.7 m/l4 L-glutamine and 40 pg/ml gentamicin. For flow cytometry, 9 X lo6 PBMC in 1.5 ml were added to 35-mm tissue-culture wells (Costar 3506, Cambridge, MA) and for immunofluorescence microscopy, 2 X lo6 PBMC in 0.5 ml were allowed to adhere onto glass coverslips placed in 16-mm tissue-culture wells (Costar 3424). After 90 min of incubation nonadherent PBMC in the above- mentioned cultures were removed and the adherent cells were washed three times with Hanks’ balanced salt solution (HBSS) (GIBCO) before fresh complete medium was added. After overnight incubation 1 pg/ml LPS (Escherichia coli, 026:B6, Sigma

LOCALIZATION OF TNF IN MONOCYTES 407

Chemical Co., St. Louis, MO) was added to the monocytes for different time periods. When monocytes were exposed to LPS for 24 hr, LPS was added on the day of cell isolation. Supernatants were collected at different time points after the addition of LPS, centrifuged (6OOg, 10 min) and frozen at -20°C until TNF activity was measured.

Flow cytometric quantijkation of TNF and TNF receptors. Monocytes cultured in 35-mm tissue-culture wells were washed once in PBS containing 2.5% pooled human A+ serum (PBS/A+) at room temperature and removed with a rubber policeman. The cells were centrifuged (6OOg, 10 min) and fixed in 2% formaldehyde (Merck, Darmstadt, West Germany) in PBS for 15 min at 0-4°C. The monocytes were permeabilized by treatment with 0.1% saponin (Sigma) in PBS for 10 min at O&C, washed once in PBS/A+, and incubated for 30 min at 37°C with 3 pg/ml polyclonal anti-TNF antibodies in PBS/A+ or a similar concentration of nRIgG. The cells were thereafter washed three times in PBS/A+ before incubation for 30 min at 37°C with FITC- conjugated goat anti-rabbit IgG F(ab’), diluted 1:20 in PBS/A+. In some experiments monocytes were stained with monoclonal antibodies against TNF (undiluted 6H 11 supernatants) followed by FITC-conjugated goat anti-mouse IgG diluted 1:20. In order to examine the presence of plasma membrane-associated TNF, nonpermeabilized (live) monocytes were stained at 0-4°C prior to fixing. To determine the specific- ity of the staining reaction, polyclonal (3 pg/ml) or monoclonal (undiluted 6Hll supernatants) antibodies against TNF were preincubated for at least 30 min with rTNF (50 pg/ml). After washing three times in PBS/A+ the cells were resuspended in PBS and analyzed using a FACScan flow cytometer (Becton-Dickinson). In some experiments live monocytes were stained with monoclonal antibodies against the human TNF receptor (10 pg/ml of htr-9b) in the same way as described above. Five thousand cells were acquired by list mode and measurements were performed on a single-cell basis and were displayed as frequency distribution histograms. Dead cells and debris were gated out of the analysis on the basis of forward light scatter.

ImmunoJluorescence microscopy. Monocytes adhered onto coverslips were washed once in PBS/A+ at room temperature before fixing and permeabilization, and then subjected to the same staining procedures as described above. The coverslips were mounted on glass slides and examined using a Zeiss Axiophot microscope (Ober- kochen, West Germany) with X 100/l .40 objective. Immunofluorescence photo- graphs were recorded on Kodak Tmax 400 film developed in Diafine (Acufine, Inc., Chicago, IL).

TNF assay. The WEHI 164 clone 13 cell line was used to quantify levels of TNF in monocyte supernatants. These cells are capable of detecting levels as low as 0.1 pg/ ml rTNF ( 10). In brief, target cells (5 X 1 O4 in 100 ~1 medium containing 0.5 pg/ml actinomycin D) were seeded into microtiter plates (Costar 3596) containing serial dilutions of the test supematants. After 20 hr of incubation cytotoxicity was measured as previously described ( 10). The amounts of TNF (pg/ml) in the test supematants were calculated on the basis of cytotoxicity obtained in the presence of various con- centrations of a rTNF standard.

RESULTS

Flow cytometric quantijication of TNF in monocytes. Affinity-purified polyclonal anti-TNF antibodies were combined with flow cytometry to quantify the production

408 HOFSLI ET AL.

of TNF in permeabilized monocytes after stimulation with LPS. Anti-TNF antibodies only weakly stained monocytes not exposed to LPS (Fig. IA). Addition of LPS resulted in an increased staining which was detectable already after 20 min (Fig. 1 B), and which peaked after 40 min (Fig. 1C). The staining intensity decreased after 60 min with LPS (Fig. 1 D), and after 120 min it was back to the level detected on unexposed monocytes (Fig. 1 E). The weak intracellular staining in nonstimulated monocytes is most likely due to activation during plastic adherence, which can induce low levels of TNF production ( 14). It is not likely that the separation media Lymphoprep could prime the monocytes for TNF production since addition of Lymphoprep to monocytes did not induce them to produce TNF. The staining of the monocytes detected with polyclonal anti-TNF antibodies was specific for TNF as incubation of the antibodies with an excess of rTNF reduced the staining to background level like that obtained with normal rabbit IgG (data not shown). Furthermore, addition of LPS did not affect the side scatter signal indicating that the size distribution of the monocytes is not influenced by LPS (data not shown). Thus, our findings indicate that LPS induces a rapid and transient increase in cell-associated TNF.

Additional experiments were carried out to quantify TNF associated with the plasma membrane of live (nonpermeabilized) monocytes. Untreated and LPS- treated monocytes were incubated with polyclonal anti-TNF antibodies, which by this treatment would have access only to TNF on the cell surface. No significant staining was detected on the plasma membrane of untreated monocytes (Fig. 2A). However, after 90 min with LPS, a small population of monocytes showed staining of the plasma membrane (Fig. 2B), and the number of positive cells increased to approximately 50% by 24 hr (Fig. 2C). The surface membrane staining observed after 24 hr with LPS probably does not represent released TNF bound back to its own receptor as: (i) untreated monocytes exposed to 10 pg/ml rTNF for 30 min at O&C did not show the same level of staining as did LPS-treated monocytes, and (ii) LPS- treated monocytes exposed to 10 pg/ml rLT for 30 min at O&C displayed high levels of membrane staining indicating that LT, which has been reported to bind to the same receptor as TNF (15), cannot displace the membrane-associated TNF. Thus, these data indicate that TNF may exist as a surface transmembrane protein in activated monocytes.

Flow cytometric quant@ication of TNF receptor expression in monocytes. Experi- ments were carried out to see if monoclonal antibodies against the TNF receptor (htr- 9b) could stain human monocytes and if such antibodies could block the membrane staining obtained with anti-TNF antibodies. The results in Fig. 3A show that htr-ob gave a specific staining of unstimulated monocytes which was blocked by preincubation of the monocytes with 50 pg/ml of rTNF. Furthermore, monocytes exposed to LPS for 24 hr showed significantly reduced staining with htr-9b as compared to unstimulated monocytes (Fig. 3B). This result could be due to binding of endogenous produced TNF to its receptor thereby blocking for htr-9b binding. It was also tested if htr-9b antibodies could affect the membrane staining obtained with anti-TNF antibodies. Ten micrograms per milliliter of htr-9b was added together with LPS for 24 hr before the monocytes were stained with anti-TNF antibodies. The results shown in Figs. 3C and 3D indicate that the antibody against the TNF receptor affects the membrane staining with anti-TNF antibodies to a minimal extent.

Zmmunojluorescence localization of TNF in monocytes. Monocytes prepared for flow cytometry were in parallel cultivated onto glass coverslips for immunofluores-


0

i ‘L nRlgG

/ / J \I

-LPS

E (42

LPS 40’

c (42:

LPS 60’

200 400 600 000 FLUORESCENCE

1000

FIG. 1. Flow cytometric quantification of intracellular TNF in monocytes. Unexposed monocytes (A) or monocytes exposed to LPS for 20 (B), 40 (C), 60 (D), or 120 (E) min were fixed, permeabilized, and stained with affinity-purified polyclonal anti-TNF antibodies (anti TNF) or normal rabbit IgG (nRIgG) as described under Materials and Methods. The numbers in brackets represent percentage cells having fluorescence corresponding to channel 280-1000. Similar data were obtained in three experiments.

410 HOFSLI ET AL.

1 - LPS

(19)

LPS 90

(22)

(

rTNF+anti TNF (8)

LPS 24 hr

0 200 400 600 600 1000 FLUORESCENCE

FIG. 2. Flow cytometric quantification of plasma membrane-associated TNF in live monocytes. Unex- posed monocytes (A) or monocytes exposed to LPS for 90 min (B) or 24 hr (C) were stained with affinity- purified polyclonal anti-TNF antibodies (anti TNF) or anti-TNF antibodies supplemented with 50 ,& ml rTNF (rTNF + anti TNF). The numbers in brackets represent percentage cells having fluorescence corresponding to channel 244- 1000. Similar data were obtained in three experiments.

cence localization of TNF. A diffuse staining of the cytoplasm was observed in permeabilized monocytes not exposed to LPS (Figs. 4A and 4B). After 20 min with LPS, perinuclear vesicles were stained in approximately all monocytes (Figs. 4C and 4D). The fluorescence intensity of the vesicles increased to a maximum at 40 min (Figs, 4E and 4F). At this time point scattered TNF-containing vesicles could be observed throughout the cytoplasm (Figs. 4E and 4F), indicating an active transport of TNF possibly to the plasma membrane. The number of monocytes displaying these distinct and bright vesicles decreased after 60 min, and at 120 min only 30-40s of the monocytes had perinuclear TNF-containing vesicles (Figs. 4G and 4H). Monoclonal antibodies against TNF (6H 11 supernatants) also stained perinuclear vesicles in LPS- stimulated monocytes (data not shown). Furthermore, the vesicular staining pattern was specific for TNF as no staining was observed when polyclonal or monoclonal anti-TNF antibodies were preincubated with an excess of rTNF (data not shown).

Experiments were also done in order to examine the presence of plasma membrane-associated TNF in live (nonpermeabilized) monocytes. Exposure of cells to


B

\ htr-Qb /

D

anti TNF anti TNF

+htr-Qb

II FLUORESCENCE (loglO)

FIG. 3. Flow cytometric quantification of TNF receptor expression in monocytes. Unexposed monocytes or monocytes exposed to LPS for 24 hr, with or without 10 &ml htr-9b, were stained with htr-9b antibodies against the TNF receptor (A, B) or with polyclonal anti-TNF antibodies (C, D). The control for htr-9b staining was monocytes exposed to 50 pg/ml of rTNF for 45 min at 0-4°C prior to staining (+TNF). Similar data were obtained in three experiments.

LPS for 24 hr resulted in a bright and punctuate staining pattern of the monocyte plasma membrane (Fig. 5A). This staining was specific for TNF as preincubation of polyclonal anti-TNF antibodies with an excess of rTNF blocked the staining (Fig. 5B).

Kinetics of TNF production in monocytes. Additional studies were performed to compare the kinetics of TNF produced at the cellular level with the kinetics of TNF release after stimulation with LPS. Supernatants from monocyte cultures prepared for flow cytometry were collected at different time points after the addition of LPS and examined for TNF activity using the WEHI 164 clone 13 bioassay. As shown in Fig. 6, TNF is released as a burst from the monocytes at 40 min of LPS exposure, whereas cell-associated TNF is detected already at 20 min and increases to a maximum at 40 min. These results clearly demonstrate that intracellular accumulation of TNF occurs before TNF is released into the medium.

DISCUSSION

The present work shows that highly specific antibodies against rTNF can be used together with flow cytometry and immunofluorescence microscopy to study the production and localization of TNF in human monocytes. The results demonstrate that LPS induces a rapid and transient production of TNF in monocytes, and that the majority of TNF is localized in distinct perinuclear vesicles which are detected before TNF is released into the medium. Furthermore, we demonstrate that flow cytometry and immunofluorescence microscopy can be used to quantify and visualize plasma membrane-associated TNF. By using other highly specific antibodies these two methods can easily be adapted to study the production and localization of other monocyte- derived cytokines.

FIG. 4. Phase contrast and immunofluorescence microscopy of intracellular TNF in monocytes. Unex- posed monocytes (A, B) or monocytes exposed to LPS for 20 (C, D), 40 (E, F), or 120 (G, H) min were fixed, permeabilized, and stained with affinity-purified polyclonal anti-TNF antibodies as described under Materials and Methods. Magnification X 1170. Similar data were obtained in three experiments.


FIG. 5. Immunofluorescence microscopy of plasma membrane-associated TNF in live monocytes exposed to LPS for 24 hr. Monocytes were stained with affinity-purified polyclonal anti-TNF antibodies (A), or with anti-TNF antibodies supplemented with 50 &ml rTNF (B). Magnification, X 1170. Similar data were obtained in three experiments.

The 233 amino acid-long TNF precursor (pre-TNF) contains an unusually long signal sequence of 76 amino acids (16). In contrast, most secretory proteins have a signal sequence of only 15-29 amino acids, which is cleaved off in the endoplasmatic reticulum ( 17). Kriegler et al. have recently suggested that the biosynthesis and secre-

1 10000

50r

30 I 1 1 4 120

II-J b”” -LPS 20 40 60 60 24hr

LPS exposure (mid

FIG. 6. Kinetics of TNF secretion (0 - 0) and accumulation (O---O) in monocytes. Monocytes exposed to LPS for different time periods were permeabilized, stained with polyclonal anti-TNF antibodies, and prepared for flow cytometry as described under Materials and Methods. Accumulation of intracellular TNF was measured as percentage cells having fluorescence corresponding to channel 280-1000 in the histograms shown in Fig. 1. In parallel, supematants of the same cultures were collected and tested for TNF activity using the WEHI I64 clone 13 bioassay. TNF levels are presented as mean rt SE of triplicate cultures. Similar data were obtained in three experiments.

414 HOFSLI ET AL.

tion of TNF are not typical for secretory proteins since pre-TNF is not cleaved until TNF is released from the cell (9). Our data show that the secretion of TNF does involve passage via intracellular vesicles which might be part of the secretory apparatus. This is in correspondence with studies by Henter et al. who found TNF present in Golgi vesicles in mononuclear cells with monocyte phenotype ( 18). Furthermore, our study shows that the localization of TNF differs from that of IL-1B which, by immunoelectron microscopy, is found in the cytosolic ground substance and not within the lumens of the secretory apparatus or on the plasma membrane of LPS- stimulated monocytes (19). Thus, the secretion of these two monokines, which are very similar in their biological activities (20), seems to occur by different mechanisms.

Earlier studies have shown that TNF activity is associated with the plasma membrane of activated monocytes (8). So far, the nature of this membrane-associated TNF is not known. Bakouche et al. concluded from their study that membrane-associated TNF represents secreted TNF bound back to its own receptor (21). On the other hand, Kriegler et al. have identified a 26-kDa cell surface transmembrane form of TNF which they suggest is clipped to generate the l7-kDa secretory component (9). In the present study we could detect plasma membrane-associated TNF on monocytes exposed to LPS for more than 90 min. The membrane-associated TNF detected after 24 hr with LPS probably does not represent released TNF bound back to its own receptor, as untreated monocytes exposed to an excess of rTNF did not show staining of the plasma membrane. This conclusion is further supported by the observation that an excess of rLT did not reduce the membrane staining of LPS- treated monocytes. It is therefore likely that the amount of TNF bound back to its receptor contributes minimally to the observed membrane-bound TNF. Thus, our results support the view that TNF may be present as a cell surface transmembrane protein. If TNF is present in secretory vesicles as a 26-kDa transmembrane protein it may appear on the plasma membrane after degranulation of the vesicles. This hy- pothesis could explain why plasma membrane-associated TNF is detected first after the peak of intracellular TNF has occurred.

It was also tested if monoclonal antibodies against the TNF receptor (htr-9b) could inhibit the membrane staining observed with anti-TNF antibodies. The htr-9b antibody blocks TNF binding on adenocarcinoma cell lines (1 I), and can also mimic the biological action of TNF in different cell systems (22). When incubated together with LPS htr-9b did not significantly reduce the anti-TNF staining of the membrane. How- ever, the htr-9b antibody does probably only recognize a part of the TNF receptor as it only partially reduces the binding of TNF to the monocytes (T. Espevik, unpublished results). It is therefore likely that multiple forms of TNF receptors exist on human monocytes.

The 17-kDa secreted form of TNF can induce most, if not all, of the multiple biological activities attributed to TNF, such as tumor necrosis (l), tumor cell killing (23), immunomodulation (2-6), septic shock (24,25), and cachexia (26). It is still not known if membrane-associated TNF can mediate the same biological activities as secreted TNF. However, earlier studies have shown that monocyte-mediated killing of some tumor cells may be mediated by membrane-associated TNF (27,28). Thus, it is tempting to speculate that the membrane-associated TNF may play an important role in regulating immune responses which require cell to cell contact.


ACKNOWLEDGMENTS

The authors thank M. R. Shalaby for critical reading of the manuscript and D. Moholdt for excellent secretarial assistance. The work was supported by grants from the Norwegian Cancer Society (Den Norske Kreftforening).

REFERENCES

1. Carswell, E. A., Old, L. J., Kassel, R. L., Green, S., Fiore, N., and Williamson, B., Proc. Natl. Acad. Sci. USA 72,3666, 191.5.

2. Scheurich, P., Thoma, B., Ucer, U., and Pfizenmaier, K., J. Immunol. 138, 1786, 1987. 3. Ranges, G. E., Zlotnik, A., Espevik, T., Dinarello, C. A., Cerami, A., and Palladino, M. A., Jr., J. Exp.

Med. 167,1472,1988. 4. Espevik, T., Figari, I. S., Ranges, G. E., and Palladino, M. A., Jr., J. Zmmunol. 140,23 12, 1988. 5. Old, L. J., Sci. Amer. 258,41, 1988. 6. Shalaby, M. R., Espevik, T., Rice, G. C., Ammann, A. J., Figari, I. S., Ranges, G. E., and Palladino,

M. A., Jr., J. Immunol. 141,499, 1988. 7. Beutler, B., Krochin, N., Milsark, I. W., Luedke, C., and Cerami, A., Science 232,977, 1986. 8. Espevik, T., and Nissen-Meyer, J., Zmmunology61,443, 1987. 9. Kriegler, M., Perez, C., DeFay, K., Albert, I., and Lu, S. D., Cell 53,45, 1988.

10. Espevik, T., and Nissen-Meyer, J., J. Immunol. Methods95,99, 1986. 11. Brockhaus, M., Loetscher, H., Hohmann, H. P., and Hunziker, W., In “2nd International Conference

on Tumor Necrosis Factor and Related Cytokines,” p. 30, Napa, CA, 1989. 12. Hohmann, H. P., Remy, R., Brockhaus, M., and van Loon, A.P.G.M., In “2nd International Confer-

ence on Tumor Necrosis Factor and Related Cytokines,” p. 3 1, Napa, CA, 1989. 13. Boyurn, A., Stand. J. Immunol. 5,9,1976. 14. Hofsli, E., Lamvik, J., and Nissen-Meyer, J., Stand. J. Zmmunol. 28,435, 1988. 15. Aggarwal, B. B., Eessalu, T. E., and Hass, P. E., Nature (London) 318,665, 1985. 16. Pennica, D., Nedwin, G. E., Hayflick, J. S., Seeburg, P. H., Derynck, R., Palladino, M. A., Kohr,

W. J., Aggarwal, B. B., and Goeddel, D. V., Nature (London) 312,724, 1984. 17. Blobel, G., Walter, P., Chang, C. N., Goldman, B. M., Erickson, A. H., and Lingappa, V. R., Symp.

Sot. Exp. Biol. 33,9, 1919. 18. Henter, J.-I., Sader, O., and Andersson, U., Eur. J. Immunof. 18,983, 1988. 19. Singer, I. I., Scott, S., Hall, G. L., Limjuco, G., Chin, J., and Schmidt, J. A., J. Exp. Med. 167, 389,

1988. 20. Le, J., and Vi&k, J., Lab. Invest. 56,234, 1987. 2 1. Bakouche, O., Ichinose, Y., Heicappell, R., Fidler, I. J., and Lachman, L. B., J. Immunol. 140, 1142,

1988. 22. Espevik, T., Brockhaus, M., and Shalaby, R., In “7th International Congress of Immunology,” Berlin,

FRG, in press. 23. Sugarman, B. J., Aggarwal, B. B., Hass, P. E., Figari, I. S., Palladino, M. A., Jr., and Shepard, H. M.,

Science 230,943, 1985. 24. Waage, A., Halstensen, A., and Espevik, T., Lancet 1,355, 1987. 25. Beutler, B., Milsark, I. W., and Cerami, A. C., Science 229,869, 1985. 26. Beutler, B., and Cerami, A., Nature (London) 320,584, 1986. 27. Espevik, T., Kildahl-Andersen, O., and Nissen-Meyer, J., Immunology57,255, 1986. 28. Kildahl-Andersen, O., Espevik, T., and N&en-Meyer, J., Cell. Immunol. 95,392, 1985.

A flow cytometric and immunofluorescence microscopic study of tumor necrosis factor production and...

Documents

Transcript of A flow cytometric and immunofluorescence microscopic study of tumor necrosis factor production and...