than lymphotoxin in vitro Tumor necrosis factor-alpha ...

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only.

1990 75: 2030-2034

Tumor necrosis factor-alpha exhibits greater proinflammatory activity than lymphotoxin in vitro CE Desch, A Dobrina, BB Aggarwal and JM Harlan

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.


Tumor Necrosis Factor-alpha Exhibits Greater Proinflammatory Activity Than Lymphotoxin In Vitro By Christopher E. Desch, Aldo Dobrina, Bharat B. Aggarwal, and John M. Harlan Tumor necrosis factor-alpha/cachectin (TNF-a) and lymphotoxin (LT, TNF-B) are primarily products of monocytes and lymphocytes, respectively. The proteins are 51% homologous in their primary structure, cause necrosis of Meth A sarcoma in vivo, are toxic t o selected tumor cells in vitro, and bind to the same receptor on cells in vitro. However, some recent studies have indicated both qualitative and quantitative differences between recombinant human (rh) LT and rhTNF with respect t o inducing human umbilical vein endothelial cell (HEC) adhesiveness for neutrophils and release of hematopoietic growth factor and interleukin-1 (IL-1) from HEC. The rhLT used in these studies was expressed in bacteria and was consequently not glycosylated, whereas natural LT is glycosylated. Therefore, we have compared various preparations of glycosy-

lated and nonglycosylated rhLT with two preparations of rhTNF with respect to their proinflammatory characteristics. We now report that the same LT cDNA, when expressed in mammaliancell line and appropriately glycosyk e d , is also markedly less potent than rhTNF on a molar basis in inducing endothelial adhesiveness for neutrophils and in directly activating neutrophil adherence to albumincoated plastic. All recombinant proteins, however, were equipotent on a molar basis in producing cytotoxicity in L929 cells. We conclude that differences in the primary structure of rhTNF and rhLT, rather than alterations induced by prokaryote protein processing, account for the disparate proinflammatory activity in vitro. 0 1990 by The American Society of Hematology.

T

binding of leukocytes.16 As a consequence of activation by TNF, endothelial cells in vitro acquire a proinflammatory and procoagulant phenotype.” Under some circumstances, TNF may also be cytotoxic to cultured endothelial cells.“ These diverse effects of T N F on endothelial cells may contribute to the tumor necrosis activity, as well as the toxicity of TNF in vivo.” Because recombinant human (rh) T N F and rhLT exhibit a similar spectrum of activities and share the same receptor on tumor cells and mammalian cells,’o~2’many investigators have considered these two proteins to be biologically equivalent. However, our previous studies demonstrated that rhTNF and rhLT expressed in Escherichia coli (rhLT-E) had divergent effects on human umbilical vein endothelial cells (HEC). RhLT was 20- to 40-fold less potent than rhTNF in inducing the release of interleukin-1 (IL-1)” and hematopoietic growth factor and the expression of adhesion molecules for leukocyte^.^^ Moreover, the binding affinity of rhLT-E for the T N F receptor on H E C was @fold less than rhTNF.’* Other investigators using the same rhLT-E have also reported significant differences between the activity of rhTNF and rhLT-E on other normal human cell types. RhTNF was more potent than rhLT in inducing osteoclastic bone re~orption,’~production of monocyte colony stimulating factor- 1 by purified human monocytes,2’ superoxide production by human neutrophils,26and release of granulocyte- and granulocyte/macrophage colony-stimulating factors by human diploid fibroblasts.’’ In contrast to our observations, Pober et a1 reported that rhTNF and rhLT are identical in their effects on HEC.” Using a glycosylated form of rhLT expressed in Chinese hamster ovary (CHO) cells, these investigators found no difference between the two molecules in the capacity to induce surface expression of E-LAM- 1 antigen or to increase surface expression of intercellular adhesion molecule- 1 and class I antigen. One possible explanation for the quantitative discrepancy between the effects of the two different rhLT preparations on HEC is the source of the cytokine. Different biologic activities could result from alterations induced by rhLT expression in E coli versus mammalian cell line (eg, glycosylation). In

H E TUMOR NECROSIS factors, TNF-a (cachectin; TNF) and lymphotoxin (TNF-8; LT), were originally defined by their ability to cause hemorrhagic necrosis in susceptible murine tumors.’ Later it was found that both cytokines were toxic to a variety of tumor cells in vitro. When recombinant proteins became available, it was soon apparent that T N F and L T also had multiple effects on normal cells, including stimulation of fibroblast proliferation,’ induction of adipocyte lipolysis,’ antiviral activity,” stimulation of bone re~orption,~ augmentation of phagocytosis and antibodydependent cellular cytotoxicity by neutrophils,6 and B cell growth-promoting activity.’ Recent studies have shown endothelial cells to be an important target of TNF. T N F induces endothelial cells to synthesize and release hematopoietic growth interleukin-1 (IL-l),’o~’’ and platelet activating factor’*; increase surface expression of class I antigen”; develop procoagulant a~tivity’~.’’; and express surface molecules that promote

From the Department of Medicine, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA: the Institute of General Pathology. University of Trieste, Trieste, Italy; the Cytokine Research Laboratory, Department of Clinical Immunology and Biological Therapy, University of Texas MD Anderson Cancer Center, Houston. TX: and the Division of Hematology, Department of Medicine. University of Washington. Seattle. WA. Submitted February 27,1989: accepted January 26,1990. Supported by Grant No. HL-03174 from the US Public Health Service. J.M.H. is a recipient of the Established Investigatorship Award from the American Heart Association. A.D. is a recipient of a Fulbright grant and a grant from the Italian Association for Cancer Research. Address reprint requests to John M. Harlan, MD. Division of Hematology ZA-34. Harborview Medical Center, 325 Ninth Ave, Seattle. W A 98104. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C.section I734 solely to indicate this fact. 0 1990 by The American Society of Hematology. 0006-4971/90/751O-OOO2$3.OO/O 2030

Blood, Vol75, No 10 (May 15). 1990: pp 2030-2034


PROINFLAMMATORY ACTIVITY

203 1

OF RHTNFV RHLT

this study we compare the effects of highly purified preparations of r h L T expressed in E coli (rhLT-E) and two r h L T preparations expressed in two different transfected mammalian cell lines (rhLT-MI, rhLT-M,) with two preparations of r h T N F expressed in E coli (rhTNF,, rhTNF,) on human endothelial cells, human neutrophils, and murine L929 cells. This comparison allows us to determine whether differences in the structure of LT account for the discrepancies between r h T N F and rhLT. Our studies demonstrate that both rhLT-M (glycosylated) and rhLT-E (nonglycosylated) are markedly less potent than r h T N F in stimulating HEC adhesiveness for neutrophils or neutrophil adhesion to albumin-coated plastic. MATERIALS AND METHODS

Reagents. RhTNF, (mol wt = 17,000 d)29and rhLT-E (mol wt = 18,000 d)30were expressed in E coli and purified to homogeneity as previously des~ribed.~'.'~ RhLT-MI was derived from the same cDNA as rhLT-E, but was expressed in a mammalian kidney cell line A293 and purified to homogeneity as described." This preparation consisted of a 20 Kd and a 25 Kd form. The 25 Kd form was identical to that previously described for the full length LT isolated from RPMI 1788.34The amino-terminal sequence of the 20 Kd polypeptide was found to be truncated by 20 residues on the amino-terminal end.-'' RhTNF, and rhLT-M, were gifts of Dr Jeffrey Browning, Biogen Inc, Cambridge, MA. The rhTNF, preparation was expressed in E coli and purified to homogeneity as previously described." RhLT-M, was purified from LT-transfected CHO cells and consisted of a major species of 20 Kd and two minor species of slightly higher molecular eight.'^ The amino-terminal amino acid sequence of this preparation was found to correspond to the amino-terminus of the 25 Kd form described by Aggarwal et al,"4 the different forms probably resulting from differences in a glycosylation. Polymyxin B and actinomycin D were purchased from Sigma Chemical Co, St Louis, MO. Cell culture. Human endothelial cells were isolated by collagenase treatment of umbilical veins as previously described" and grown in RPMI-1640 medium (Whittaker MA Bioproducts, Walkersville, MD) containing 10 mmol/L N'-2-hydroxyethylpiperazineN'-ethanesulfonic acid, 20% adult bovine and newborn calf serum (1:l; Hyclone Sterile Systems, Logan, UT), 20 pg/mL endothelial cell-derived growth factor prepared as described by Maciag et al," and 90 pg/mL heparin as described by Thornton et aLJ9 Adherence assays. Neutrophil adherence to HEC monolayers was assayed as previously described.I6 Human polymorphonuclear neutrophils (PMNs) were purified by Ficoll-Hypaque gradient centrifugation, dextran sedimentation, and hypotonic lysis of contaminating red blood cells. The PMNs were labeled with "Cr (as sodium chromate, 1 pCi/mL; New England Nuclear, Boston, MA) as described.40The "Cr-labeled PMNs were resuspended in RPMI1640 medium at 2,00O/pL. First- or second-passage HEC were plated at confluent density in 48-well plates. Growth medium was removed, and RPMI-1640 medium (250 pL) containing 2% normal calf serum, 10 pg/mL polymyxin B, and varying concentrations of the cytokines was added. After a 4-hour incubation at 37"C, the HEC monolayers were washed once with phosphate-buffered saline (PBS), and 250 pL of medium containing 5 x IO5 "Cr-labeled PMNs was added. After a 30 minute incubation, supernatant medium and nonadherent PMNs were decanted, and each well was washed once with PBS. Adherent 5'Cr-labeled PMNs were solubilized with 500 pL of 1 mol/L ammonium hydroxide, and the entire material from each well was analyzed for radioactivity in a gamma

counter. Adherence was calculated using the following equation: 5% adherence = (cpm "Cr in NH40H lysate/total cpm "Cr added) x 100. PMN adherence to albumin-coated plastic wells was determined as follows. PMN were isolated and labeled with "Cr as described above, and suspended in RPMI-1640 containing 2% normal calf serum at a concentration of 2,0OO/pL. Forty-eight-well tissue culture plastic plates were incubated for 1 hour with 250 pL per well of PBS containing 1% heat-treated (8OOC for 3 minutes) bovine serum albumin (BSA, Irvine Scientific, Santa Ana, CA). The wells were washed twice with PBS before the addition of 51Cr-labeled PMN. Medium containing "Cr-labeled PMN (250 pL) was then added to each well. Test reagents were diluted in RPMI containing 2% calf serum and 10 pg/mL polymyxin B. Medium alone or medium containing the test reagent (250 pL) was added to the wells immediately after addition of PMNs. The PMNs were allowed to adhere to BSA-coated plastic for 1 hour. The plates were then washed twice with PBS to remove nonadherent PMNs. Adherent "Cr-labeled PMNs were harvested with NH40H, and percent adherence was calculated as above. L929 cytotoxicity assay. The cytotoxic activity of rhTNF, rhLT-E, and rhLT-M on murine L929 cells was determined using Briefly, L929 murine actinomycin D as a sensitizing agent!' fibroblasts (ATCC CCL 1, American Type; Culture Collection, Rockville, MD) were grown in 75 cm2 flasks in RPMI-1640 containing 10%horse serum (Flow Laboratories, McLean, VA). The cells were harvested by trypsinization, suspended in modified minimal essential medium (Flow Laboratories) containing 10% horse serum, and plated in microtiter wells at 2.5 x IO4 cells per well. At the time of the assay, actinomycin D (1 pg/mL final concentration) was added to each well followed by control medium or medium containing rhTNF, rhLT-E, or rhLT-M. After an overnight incubation, supernatant medium was decanted, and the wells were washed three times with PBS. The remaining adherent cells were fixed and stained with crystal violet in 20% methanol for 15 minutes, and then washed again four times with water to remove unbound dye. Stained cells were lysed with 0.1 mol/L sodium citrate pH 4.2/50% ethanol, and the plates were read in a photometer at an absorbance of 570 nm (As7,). Control wells were incubated in medium containing actinomycin D only (A570medium). Maximum cytotoxicity was determined by incubation wtih 100 U/mL of freshly diluted rhTNF (A,,, maximum). Cytotoxicity was calculated as follows: % Cytotoxicity = 1 -

A570 Test - A570 Maximum x 100% A570 Medium - A570 Maximum

Statistics. Data were analyzed by analysis of variance treatment (GB-Stats; Dynamic Microsystems, Inc, Silver Springs, MD). RESULTS

On a molar basis rhTNF,, rhLT-E, and rhLT-MI produced equivalent cytotoxicity to murine L929 cells (Fig 1). Fifty percent cytotoxicity was demonstrated at approximately 6 pmol/L rhTNF,, 4 pmol/L rhLT-E, and 8 pmol/L rhLT-M, (means of five separate experiments). Concentrations of each recombinant protein greater than 50 pmol/L produced 100%cytotoxicity. Each recombinant protein was tested for its ability to induce HEC to express surface activity promoting PMN adherence. Treatment of HEC with rhTNF,, rhLT-E, or rhLT-MI for 4 hours resulted in a dose-dependent increase in PMN adherence (Fig 2). R h T N F , was markedly more potent than either rhLT-E or rhLT-MI in inducing HEC adhesiveness for PMNs. PMN adherence to HEC pretreated

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. 2032

DESCH ET AL

100,

25

T

I

0

I

10

20

30

25

50

DM PM

Fig 1. Cytotoxic activity of rhTNF,. rhLT-E. and rhLT-M,. Cytotoxicity was determined after an 18-hour incubation of actinomycin D-treated L929 cells with rhTNF, (0).rhLT-E (O),and rhLT-M, (A). Values represent the means f SEM of five separate experiments. Values for rhTNF, were not significantly different from those for rhLT-E and rhLT-M, by analysis of variance.

with 250 pmol/L rhTNF, was 20% * 1% versus 9% 2 3% with 500 pmol/L rhLT-E, and 6% 2 2% with 500 pmol/L rhLT-MI (means * SEM of five experiments). Stimulation of HEC adhesiveness for PMNs was observed at rhTNF, concentrations as low as 15 pmol/L, whereas significant stimulation of HEC was detected only at rhLT-E or rhLTMI concentrations greater than 60 pmol/L. The marked disparity between rhTNF and rhLT was confirmed in a separate series of experiments with additional preparations of rhLT-M and rhTNF. RhLT-M, was derived from CHO cells and, therefore, also glycosylated. RhLT-M, was markedly less active than rhTNF, in stimulating endothelial adhesiveness for PMNs (Fig 3), although both produced equivalent cytotoxicity on L929 cells. PMN adherence to HEC pretreated with 25 pmol/L rhTNF, was 17% r 3% versus 7% 2 3% to HEC pretreated with 50 pmol/L rhLT-M, (means i. SEM of three experiments). In the L929

Fig 3. Effect of rhTNF, and rhLT-M, on HEC adhesiveness for neutrophils. Neutrophil adherence was determined after a 4-h incubation with medium alone 13% f 1%). rhTNF, IO), or rhLT-M, (0).Values represent the means f SEM of three separate experiments. Values for rhTNF, were significantly different from those for rhLT-M, by analysis of variance (P< .OOOl).

assay, 50% cytotoxicity was observed with 2 pmol/L rhTNF, and 1 pmol/L rhLT-M,. RhTNF,, rhLT-E, and rhLT-MI were also tested for their capacity to stimulate directly human PMN adherence to BSA-coated plastic (Fig 4). Again, rhTNF, was significantly more potent than rhLT-E and rhLT-MI at each concentration tested. Stimulated adherence was 40% 2 5% with 10 nmol/L rhTNF, versus 28% 2 3% and 21% 2 2% with 10 nmol/L rhLT-E and rhLT-M,. At 0.1 nmol/L, adherence was 24% 2 4% with rhTNF, versus 11% 2 3% and 12% 2 3% with rhLT-E and rhLT-M,, respectively (means 2 S E of seven experiments). DISCUSSION

Although both rhTNF and rhLT activate HEC and PMNs, our studies demonstrate that rhTNF is markedly more potent than rhLT, regardless of whether rhLT is

25 T

50 T

1

ti

o/-0

125

250

I

I

375

500 "

PM

I

2.5

5.0

7.5

1c

nM

Fig 2. Effect of rhTNF,, rhLT-E. and rhLT-M, on HEC adhesiveness for neutrophils. Neutrophil adherence was determined after a Qhour incubation of HEC with medium alone (3% + 2%). rhTNF, (01, rhLT-E (01, or rhLT-M, ( A ) . Values represent the means f SEM of five separate experiments. Values for rhTNF, were significantly different from those for rhLT-E or rhLT-M, by analysis of variance (P< .OOOl).

Fig 4. Effect of rhTNF,, rhLT-E. and rhLT-M, on neutrophil adherence t o BSA-coated plastic. Neutrophil adherence t o BSAcoated plastic was determined after a 1-hour incubation with medium alone (8% f 4%). rhTNF, (0).rhLT-E (01,or rhLT-M, ( A ) . Values for rhTNF, were significantly different from those for rhLT-E or rhLT-M, by analysis of variance (P< .OOOl).


2033

PROINFLAMMATORY ACTIVITY OF RHTNF v RHLT

derived from E coli (rhLT-E) or a mammalian cell source (rhLT-M). The disparity between r h T N F and rhLT-M is in agreement with previously published differences between r h T N F and rhLT-E.2'-26 O u r results with two well-characterized preparations of rhLT-M effectively eliminate the possibility that glycosylation or other differences between prokaryote and eukaryote protein processing account for the previously reported disparity between the activity of rhLT-E versus r h T N F on HEC.22,23 Pober et a12*reported that rhTNF expressed in E coli and rhLT expressed in CHO cells were equivalent in their ability to activate HEC. Unfortunately, the amino-terminal amino acid sequence of the rhLT preparation used in their studies has not been published, and the discrepancy between their results and ours remains unexplained. In summary, although rhTNF and rhLT are remarkably similar in a number of respects, including tumor necrosis activity in vivo, cytotoxicity for some tumor cells in vitro, and binding to the same receptor on some cell types in vitro, the proteins differ significantly in their biologic effects on human

endothelial ~ e l l s ; ~human , ~ ~ m0nocytes,2~human neutrophils;6 human fibroblast^,^' and several human tumor cells36in vitro. The significant differences between the effects of rhTNF and rhLT on human cells in vitro warrant a critical comparison of the activities of the two proteins in vivo. Since rhLT-E and rhTNF produce similar effects in mice,"2 it may be necessary to compare the proteins in a nonhuman primate. If rhLT is indeed less potent than rhTNF in activating monocytes, neutrophils, fibroblasts, and endothelial cells in vivo, it will likely be considerably less toxic than rhTNF. Such studies may also provide insight into whether "toxic" effects are necessary for the anti-tumor activity of the tumor necrosis factors. ACKNOWLEDGMENT

The skilled technical support of Kathe Stanness and Penny Thompson is gratefully acknowledged. We thank Pauline Marsden for her skillful word processing and Dr Nicholas Kovach for performing the statistical analysis.

REFERENCES 1. Aiyer RA, Aggarwal BB: Tumor necrosis factors, in Podack

ER (ed): Cytolytic Lymphocytes and Complement Effectors of the Immune System, vol 11. Boca Raton, FL, CRC, 1988, p 105 2. Sugarman BJ, Aggarwal BB, Hass PE, Figari IS, Palladino MA, Shepard HM Jr: Recombinant human tumor necrosis factoralpha: Effects on proliferation of normal and transformed cells in vitro. Science 230:943, 1985 3. Patton JS, Shepard HM, Wilking H, Lewis G, Aggarwal BB, Eessalu TE, Gavin LA, Grunfeld C: Interferons and tumor necrosis factors have similar catabolic effects on 3T3 L1 cells. Proc Natl Acad Sci USA 83:8313,1986 4. Wong GHW, Goeddel DV: Tumour necrosis factors alpha and beta inhibit virus replication and synergize with interferons. Nature 323:819, 1986 5. Bertolini DR, Nedwin GE, Bringman TS, Smith DD, Mundy GR: Stimulation of bone resorption and inhibition of bone formation in vitro by human tumour necrosis factors. Nature 319516,1986 6. Shalaby MR, Aggarwal BB, Rinderknecht E, Svedersky LP, Finkle BS, Palladino MA Jr: Activation of human polymorphonuclear neutrophil functions by interferon-gamma and tumor necrosis factors. J Immunol 135:2069, 1985 7. Kehrl JH, Alvarez-Mon M, Delsing GA, Fauci AS: Lymphotoxin is an important T cell-derived growth factor for human B cells. Science 238:1144, 1987 8. Munker R, Gasson J, Ogawa M, Koeffler H P Recombinant human TNF induces production of granulocyte-monocyte colonystimulating factor. Nature 323:79, 1986 9. Broudy VC, Kaushansky K, Segal GM, Harlan JM, Adamson JW: Tumor necrosis factor-alpha stimulates human endothelial cells to produce multilineage hematopoietic growth factor(s). Proc Natl Acad Sci USA 83:7467,1986 10. Nawroth PP, Bank I, Handley D, Cassimeris J, Chess L, Stern D: Tumor necrosis factor/cachectin interacts with endothelial cell receptors to induce release of interleukin 1. J Exp Med 163:1363, 1986 1 1. Libby P, Ordovas JM, Auger KR, Robbins AJ, Birinyi LK, Dinarello CA: Endotoxin and tumor necrosis factor induce interleukin-1 gene expression in adult human vascular endothelial cells. Am J Pathol 124:179, 1986

12. Bussolini F, Camussi G, Baglioni C: Synthesis and release of platelet-activating factor by human vascular endothelial cells treated with tumor necrosis factor or interleukin-1-alpha. J Biol Chem 263:11856,1988 13. Collins T, Lapierre LA, Fiers W, Strominger JL, Pober JS: Recombinant human tumor necrosis factor increases mRNA levels and surface expression of HLA-A, B antigens in vascular endothelial cells and dermal fibroblasts in vitro. Proc Natl Acad Sci USA 83:446, 1986 14. Nawroth PP, Stern DM: Modulation of endothelial cell hemostatic properties by tumor necrosis factor. J Exp Med 164:740, 1986 15. Bevilacqua MP, Pober JS, Majeau GR, Fiers W, Cotran RS, Gimbrone MA Jr: Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: Characterization and comparison with the actions of interleukin 1. Proc Natl Acad Sci USA 83:4533,1986 16. Gamble JR, Harlan JM, Klebanoff SJ, Vadas MA: Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc Natl Acad Sci USA 82:8667, 1985 17. Bevilacqua MP, Gimbrone MA Jr: Inducible endothelial functions in inflammation and coagulation. Sem Thromb Hemost 13:425, 1987 18. Sato N, Goto T, Haranaka K, Satomi N, Hariuchi H, Mano-Hirano Y, Sawasaki Y: Actions of tumor necrosis factor on cultured vascular endothelial cells: Morphologic modulation, growth inhibition, and cytotoxicity. J Natl Cancer Inst 76:1113, 1986 19. Have11 EA, Fiers W, North RJ: The antitumor function of tumor necrosis factor (TNF). I. Therapeutic action of TNF against an established murine sarcoma is indirect, immunologically dependent, and limited by severe toxicity. J Exp Med 167:1067, 1988 20. Aggarwal BB, Eessalu TE, Hass PE: Characterization of receptors for human tumour necrosis factor and their regulation by gamma-interferon. Nature 318555, 1985 21. Stauber GB, Aggarwal BB: Characterization and affinity cross linking of receptors for human recombinant lymphotoxin (tumor necrosis factor-@ on a human histiocytic lymphoma cell line, U-937. J Riol Chem 264:3573, 1989

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. 2034 22. Locksley RM, Heinzel FP, Shepard HM, Agosti J, Eessalu TE, Aggarwal BB, Harlan JM: Tumor necrosis factors alpha and beta differ in their capacities to generate interleukin 1 release from human endothelial cells. J Immunol 139:1891, 1987 23. Broudy VC, Harlan JM, Adamson JW: Disparate effects of tumor necrosis-factor-alpha/cachectin and tumor necrosis factorbeta/lymphotoxin on hematopoietic growth factor production and neutrophil adhesion molecule expression by cultured human endothelial cells. J Immunol 138:4298, 1987 24. Thomson BM, Mundy GR, Chambers TJ: Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J Immunol 138:775,1987 25. Oster W, Lindemann A, Horn S, Mertelsmann R, Hermann F Tumor necrosis factor (TNF)-alpha but not TNF-beta induces secretion of colony stimulating factor for macrophages (CSF-I) by human monocytes. Blood 70:1700, 1987 26. Figari IS, Mori NA, Palladino MA Jr: Regulation of neutrophil migration and superoxide production by recombinant tumor necrosis factors-alpha and -beta: Comparison to recombinant interferon-gamma and interleukin-1 alpha. Blood 70:979, 1987 27. Koeffler HP, Gasson J, Ranyard J, Souza L, Shepard M, Munker R: Recombinant human T N F alpha stimulates production of granulocyte colony-stimulating factor. Blood 7055, 1987 28. Pober JS, Lapierre LA, Stolpen AH, Brock TA, Springer TA, Fiers W, Bevilacqua MP, Mendrick DL, Gimbrone MA Jr: Activation of cultured human endothelial cells by recombinant lymphotoxin: Comparison with tumor necrosis factor and interleukin 1 species. J Immunol 138:3319,1987 29. Pennica D, Nedwin GE, Hayflick JS, Seeburg Ph, Derynck R, Palladino MA, Kohr WJ, Aggarwal BB, Goeddel DV: Human tumour necrosis factor: Precursor structure, expression and homology to lymphotoxin. Nature 312:724, 1984 30. Gray PW, Aggarwal BB, Benton CV, Bringman TS, Henzel WJ, Jarrett JA, Leung DW, Moffat B, Ng P, Svedersky LP, Palladino MA, Nedwin G E Cloning and expression of cDNA for human lymphotoxin, a lymphokine with tumour necrosis activity. Nature 312:721, 1984 31. Aggarwal BB, Kohr WJ, Hass PE, Moffat B, Spencer SA, Henzel WJ, Bringman TS, Nedwin GE, Goeddel DV, Harkins RN:

DESCH ET AL

Human tumor necrosis factor: Production, purification, and characterization. J Biol Chem 2602345, 1985 32. Aggarwal BB, Moffat B, Harkins RN: Human lymphotoxin: Production by a lymphoblastoid cell line, purification and initial characterization. J Biol Chem 259:686, 1984 33. Hains JM, Aggarwal BB: Characterization of recombinant human lymphotoxin (TNF-j3) produced by a mammalian cell line. Arch Biochem Biophys 274:417,1989 34. Aggarwal BB, Henzel WJ, Moffat B, Kohr WJ, Harkins RN: Primary structure of human lymphotoxin derived from 1788 lymphoblastoid cell line. J Biol Chem 260:2334, 1985 35. Marmenout A, Franson L, Travernier J, van der Heyden J, Tizard R, Kawashima E, Shaw A, Johnson M-J, Semon D, Mueller R, Ruysschaert M-R, van Vlhet A, Fiers W: Molecular cloning and expression of human tumor necrosis factor and comparison with mouse tumor necrosis factor. Eur J Biochem 152515, 1985 36. Browning J, Ribolini A: Studies on the differing effects of tumor necrosis factor and lymphotoxin on the growth of several human tumor lines. J Immunol 143:1859,1989 37. Jaffe EA, Nachman RL, Becker CG, Minick C R Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest 52:2745, 1973 38. Maciag T, Cerudolo J, Ilsley S, Kelly PR, Forand R: An endothelial cell growth factor from bovine hypothalamus: Identification and partial characterization. Proc Natl Acad Sci USA 765674, 1979 39. Thornton SC, Mueller SN, Levine EM: Human endothelial cells: Use of heparin in cloning and long-term serial cultivation. Science 222:623, 1983 40. Gallin JT, Clark RA, Kimball HR: Granulocyte chemotaxis: An improved in vitro assay employing "Cr-labelled granulocytes. J Immunol110:233,1973 41. Flick DA, Gifford GE: Comparison of in vitro cell cytotoxic assays for tumor necrosis factor. J Immunol Methods 68:167,1984 42. Kaushansky K, Broudy VC, Harlan JM, Adamson JW: Tumor necrosis factor-alphaand tumor necrosis factor-beta (lymphotoxin) stimulate the production of granulocyte-macrophage colonystimulating factor, macrophage colony-stimulating factor and interleukin-1 in vivo. J Immunol 141:3410, 1988