TNF Receptors Irnmunol Res 1992;11:81-90
Joachim Rothe G&ela Gehr Hansruedi Loetscher Werner Lesslauer
Tumor NecrosisFactor ReceptorsStructure and Function
F. Hoffmann-LaRoche, Ltd., Basel, Switzerland
oooo*loo~176176176176176176
tooo*eeoooel.*..~176176176176176176
Key Words
Abstract
Tumor necrosis factor, receptors and inhibitory proteins Nerve growth factor receptor family Cytokines
Tumor necrosis factors (TNFs) have been a focus of research for well over a decade now. The identification and recent molecular cloning of two different types of ceil-surface TNF receptors will shed further light on the mode of action of these pleiotropic cytokines. In the present article, we summarize the data on the biochemistry and structure of the receptors and focus on the molecular cloning of the respective cDNAs. The nucleotide sequences of the receptor genes revealed that both TNF receptors belong to the still growing nerve growth factor receptor gene family. The function and origin of TNF inhibitory proteins as well as receptor-mediated signal transduction are discussed.
JeloooooooooQo*ooo~176176176176176176
* e e e e e e e o * e e e e e e e
The functional activities of cells in immunological and inflammatory host defense reactions are mediated by a large and still growing number of protein factors such as interferons, interleukins, colony-stimulating factors and tumor necrosis factors (TNF), which are commonly referred to as cytokines. Cytokines act by binding to specific cell surface receptors. Their mode of action in general is local and paracrine rather than sys-
temic, and their serum concentrations usually are very low. Typically, a cytokine is produced by several different cell types and its receptors are expressed on a variety of potential target cells. It may elicit the production of secondary cytokines in a cascade-like fashion or modulate cytokine receptor function. Specific mechanisms such as the controlled release of cytokine inhibitors limit cytokine activities. A complex regulatory network of me-
Dr. W. Lesslauer Building 15, Room 6 F. Hoffmann-La Roche Ltd. -CH-4002 Basel (Switzerland)
9 1992 S. Karger AG, Basel 0257-277X/92/ 0112-0081 $2.75/0
diators thus is generated which forms the basis of the pleiotropic and redundant cytokine functions. TNF was first discovered by its ability to induce hemorrhagic necrosis of certain tumors [ l, 2]. Later, after the molecular cloning and recombinant expression, a wide range of biological activities was discovered. TNF is now considered to be one of the most multifunctional response modifiers in inflammatory processes and immunological reactions, to have growth factor activity on some cell types and to induce expression of adhesiontype or histocompatibility molecules on others [for reviews, see ref. 3, 4]. In some instances, extremely harmful effects are produced by TNF. For example, the lethal consequences of septic shock are largely mediated by TNF [5]. Furthermore, tissue damage in bacterial meningitis correlates with TNF concentration in cerebrospinal fluid [6]. The term TNF commonly refers to the two related factors, TNFa (cachectin) and TNF~ (lymphotoxin), which have a sequence homology of about 30% [7]. TNFa and TNF[3 are thought to be mainly produced by monocytes/macrophages and activated T lymphocytes. The three-dimensional structure of TNFa has been investigated in X-ray diffraction studies [8, 9]. TNFa crystallizes as a trimer, and the biologically active forms of both TNFa and TNFI3 are believed to be trimers [ 10]. Recently, several groups identified two distinct TNF receptors (TNFR), termed TNFRa (TNFR type II, p75, utr antigen) and TNFR[3 (TNFR type I, p55, htr antigen) in man [1120] and mouse [21, 22]. When the predicted amino acid sequences of the two receptors were analyzed, a high degree of similarity of the extracellular regions emerged, whereas the sequences of the intracellular regions were entirely unrelated. These findings raised the questions of whether the two receptors have distinct functional properties, and how the
82
many different activities of TNF relate to TNFRct and TNFRI3. The present review focuses on the structure of the two receptors and summarises current views of their biological activities.
Protein Biochemistry of TNFR In the past few years TNFR have been identified on the surface of most human and murine cells analysed in a large number of studies [for references, see ref. 23]. In radioligand binding studies, typically a single class of cellular TNF binding sites with Ka values in the picomolar range was found. Chemical crosslinking of radioiodinated TNF to cell surface binding sites led to the identification of various TNFa- and TNFI3-reactive bands in SDS-PAGE from about 50 to 138 kD. The analysis of the crosslinked complexes from various cells indicated at least two distinct TNFR molecules which differ not only in molecular mass but also in ligand-binding affinity, glycosylation, immunoreactivity and proteolytic fingerprints [24]. TNFRa has a molecular mass of 75 kD, TNFR[3 is smaller (55 kD). The existence of two different receptors was confirmed by monoclonal antibodies raised against partially purified TNFR fractions; two classes of antibodies were found which reacted exclusively with either TNFRa or TNFR[3 [25]. The receptors are expressed in different relative amounts on various cell types in the range of several hundred to a few thousand molecules per cell. The two TNFR were purified from the HL60 and U937 cell lines, and from human placenta by combined immuno- and ligandaffinity column chromatography and reversephase high performance liquid chromatography [23, 26, 27]. Both receptors were found to bind TNFa and TNF[3 with Ka values in the picomolar range, indicating that no accessory
Rothe/Gehr/Loetscher/Les slauer
TNF Receptors
protein is necessary to form a high-affinity ligand-binding site with either TNFR [10]. The receptor proteins purified from HL60 cells revealed major bands at 75 kD for TNFRa and 55 kD for TNFRI] in addition to a few minor bands (see below) by SDS-PAGE analysis. Specific binding of 12SI-TNFa to all these bands was demonstrated in ligand blot experiments [11]. Furthermore, the 75- and 55-kD bands reacted exclusively with monoclonal antibodies specific for TNFRa and TNFRI3, respectively [11, 25]. Partial N-terminal and internal peptide sequences were determined for the 75kD and 55kD bands. Protein sequencing of the minor bands which copurified with TNFRct and TNFR~ revealed that they were derivatives or fragments of the receptors [13, 26]. Interestingly, a somewhat shortened form of TNFRa was found to be ubiquitinated; the functional significance of the ubiquitination remains to be established.
Molecular Cloning of the Two TNFR
Several groups have reported the molecular cloning of human [12-20] and mouse TNFRct and TNFRJ3 [21, 22]. In our own work we have isolated the cDNAs encoding human TNFRa and TNFR[3 using partial amino acid sequence information and a combination of polymerase chain reaction and conventional cloning strategies [12, 13]. The mouse TNFR cDNAs were isolated from murine cDNA libraries by interspecies crosshybridization using a fragment of the homologous human cDNA as probes [21, 22]. The open reading frame of the human TNFRJ3 cDNA encodes a protein of 455 amino acids with a typical signal peptide of 29 amino acids and a single 21-amino acid transmembrane region, which separates the extracellular Nterminal domain (182 residues including 24 cysteines) from the C-terminal cytoplasmic
domain (223 residues). Three potential Nlinked glycosylation sites are contained in the extracellular domain. The amino terminal region of the molecule is extracellularly located, because the recombinant N-terminal region binds TNFet and antibodies which bind to intact cells. Leucine was determined to be the N-terminal residue by protein sequencing. Mature human TNFRI3 therefore consists of 426 amino acids with a theoretical molecular mass of 47.5 kD. The human TNFR~ cDNA encodes a receptor protein of 439 amino acids with a single transmembrane region and a theoretical molecular mass of nearly 50 kD. The extracellular and intracellular regions are composed of 235 and 178 residues, respectively. Two potential N-linked glycosylation sites are present in the extracellular domain. Posttranslational modifications of both receptors account for the differences between the theoretical and the apparent molecular weights estimated from SDS-PAGE. A hypothetical overall structure of both receptor molecules is schematically illustrated in figure I. Transfection of either of the two human TNFR cDNAs into COS-1 cells led to the transient expression of a single class of highaffinity TNFa binding sites. Apparent Kd were determined with the transfected COS-1 cells by radioligand binding assays; values of about 0.1 nM(TNFRa) and 0.5 nM(TNFRI3) were found [ 12, 13]. The predicted amino acid sequences of the two mouse TNFR reveal high degrees of similarity to the homologous human receptors with regard to overall size, relative size of domains and potential glycosylation pattern. The high degree of sequence conservation between the human and mouse TNFR is most clearly demonstrated by the fact that, as shown in table 1, the amino acid sequence identity in the extracellular domains of human and mouse TNFRct and TNFR[3 reaches 58 and 70 %, respectively [ 12-15, 21, 22].
83
TNFR Belong to a New Receptor Gene Family
ct M,H
i I
ii II H M,H
iii fll
H M,H
iv
IV
Cytoplasm
Fig. 1. Schematic representation of TNFRa and TNFR!3. Cysteine-riche repeats are designated by Roman numbers, conserved cysteine residues itself by horizontal lines inside the repeats. Circles are potential N-linked glycosylation sites in man (H) and mouse (M).
Table 1. Percent amino acid identity between corresponding parts of TNF receptors in man (h) and mouse (m)
Leader EC domain TM region IC domain
hTNFRttmTNFRct
hTNFRI]mTNFRI3
TNFRctTNFRI3
55 58 65 73
76 70 74 59
19 22 28 < 10
Values given in the last section are approximate numbers. EC = Extracellular, TM = transmembrane and IC = intracellular,
84
The predicted amino acid sequences of the extracellular regions of the two TNFR in man and mouse reveal significant similarities. The most prominent feature is a distinct pattern of cysteine residues defining four conserved sequence repeats of roughly forty amino acids shown schematically in figure t. The two N-terminal repeats of both receptors contain six cysteines with a consensus pattern ofCX]0_~ sCX0_2CXaCXs_1 ICX3-8C. The third and fourth repeats contain six (TNFR[~) or four cysteines (TNFRa) in homologous positions and thus are less well conserved between the receptors. The alignment, however, still scores significantly above a random score with the mutation data matrix. Five other currently known sequences have striking similarities with the TNFR extracellular regions: the human and rat low-affinity nerve growth factor receptors, the B cell surface antigen CD40, the murine cDNA clone 4-1 BB isolated from induced helper and cytolytic T cell clones, the transcriptionally open reading frame T2 from the Shope fibroma virus, and the OX 40 antigen present on activated CD4-positive rat Tlymphocytes [for references, see ref. 28]. The similarity among all these receptor-type molecules, however, is restricted to the presumably heavily disulfide-bonded extracellular regions. Interestingly, the level of sequence identity between the two TNFR is lower than that to other members of the receptor gene family. The evolutionary conservation of the cysteine-rich repeat motif suggests that this sequence folds into a three-dimensional scaffold, which is favorable for the binding of protein ligands. The fine specificity for binding a specific ligand according to this hypothesis is determined by the intervening amino acids. Electrostatic interactions may play a role. It is noteworthy that both TNFR carry an overall
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
positive charge at physiologic pH, while TNF is negatively charged. In contrast, the negatively charged extracellular domain of the NGFR is complementary to the positively charged NGF. The intracellular regions of the two TNFR do not show any obvious sequence similarity to each other nor to any other known mammalian sequence. In particular, no similarities to catalytic domains of tyrosine- or serine/ threonine-specific protein kinases are present. There are some patterns which resemble sequence elements involved in signal transduction in other molecules, i.e. potential phosphorylation sites for tyrosine kinase (TNFRI3), protein kinase C (TNFRct and TNFRI3) and cyclic nucleotide dependent kinase (TNFRct and TNFR[3). Interestingly, the cytoplasmic domains of both TNF receptors have a high content in serine, threonine and proline. These so-called PEST sequences might relate to high turnover rates and intracellular protein degradation [29].
TNF Inhibitory Proteins A few years ago, two proteins were discovered in human serum and urine which inhibited TNF activity in cell culture. They were termed TNF-BP I and II and shown to act by binding to TNFct and, at least for TNF-BP I, to a lesser extent to TNFI3 [30-35]. When partial amino acid sequences of the TNF-BP had been determined, both TNF-BP were recognized as truncated fragments of the extracellular regions of the respective TNFR [35-37]. The serum concentrations of the TNF-BP have not yet been systematically investigated, but in one report they have been found to occur at concentrations in the range of ng/ml [38]. There is general agreement that normal TNFct concentrations in serum are about 30 pg/ml or lower. TNF-BP are therefore present
in large molar excess, but their physiologic role is not fully understood. They may act by neutralization of TNF in the systemic circulation, thus enforcing paracrine type activity. Alternatively, they may be viewed as a buffer from which TNF is slowly released. The view that they exert a TNF-neutralizing activity is supported by animal studies in which natural TNF-BP [39] or recombinant TNFR molecules [40] have been found to inhibit TNFmediated pathology.
Recombinant Soluble TNFR The extracellular regions of both TNFR have been expressed in recombinant soluble form (rec sol TNFR) in the baculovirus/Sf9 or CHO cell expression systems and the TNF binding properties were determined [41]. With rec sol TNFRI3, a single class of highaffinity binding sites for TNFct with K,~ values between 0.3 and 0.4 ru'v/was found. In contrast to the native cell surface-bound receptor which does not distinguish between TNFct and TNF[3 [24, 42], rec sol TNFRI3 showed a significantly lower affinity for TNFI3 compared to that for TNFcc The Kd values ranged from 0.5 to 1.6 nM depending on the expression system used. This is consistent with the previously reported finding that the naturally occurring soluble TNF-BP I is less potent in neutralizing TNFI3 [31, 35]. Interestingly, the Scatchard analysis of TNFct binding to rec sol TNFRct in a solidphase binding assay revealed a nonlinear curve which suggested two affinity classes for ligand binding [Loetscher et al., unpubl, data]. The best fit to the binding curve was achieved with Kd of 0.2 nM (high affinity site) and 2.5 nM (low affinity site). The analogous Kd values for the binding of TNFfl to rec sol TNFRct were in the range of 0.2-0.6 and 3.7-10 nM for the high and low affinity sites, respective-
85
ly. The significance of the low-affinity TNF binding site of rec sol TNFRa remains to be established. It is noteworthy that a single class of binding sites of membrane-bound TNFRct has been reported in a number of studies. It might be proposed that the low-affinity site is generated by a conformational change induced by the free C-terminal end of rec sol TNFRa. However, both high and low affinity TNFa binding sites with Kd values similar to those described above were recently reported for both native and recombinant full-length TNFRct [15]. Chimeric proteins in which the extracellular domains of human TNFRa and TNFRI3 were fused to the hinge region of human IgG y3 (rsTNFR~t-h73 and rsTNFRI3-hy3) have been expressed in mouse myeloma cells [41]. These fusion proteins have a disulfidebonded dimeric structure similar to that of an antibody molecule. In ligand-binding studies it was found that the fusion proteins bind TNFct and TNFI3 with high affinity. A linear relationship was obtained in the Scatchard analysis with Ka values between 0.1 and 0.2 n_M for both TNFct and TNFI3. The various rec sol TNFR constructs were found to compete with full-length TNFRct and TNFR[3 for TNF binding in in vitro binding assays. Furthermore, it could be shown that the cellular cytotoxicity of TNF in cell culture is inhibited by rec sol TNFR in a concentration-dependent fashion. The comparison of the TNFneutralizing activity demonstrated that rsTNFRf3-h73 inhibits TNF bioactivity at 10to 100-fold lower concentrations than rsTNFRI3. The molecular mass of rsTNFRI3 in solution as determined by light scattering and analytical ultracentrifugation studies is about 25 kD [41]. Stable complexes of rsTNFRI3 with TNFa and TNFI3 were formed and were both found to have a molecular mass of about 140 kD, suggesting a stoichiometry of three
86
rsTNFRI3 bound to one TNF trimer. The stability of these complexes is of interest since microclustering of TNFRI3 has been proposed as an essential step in TNF signal generation at the surface of target cells [43].
Functional Aspects The discovery of two distinct TNFR raised the question whether they have equivalent or distinct functions. It was soon realized, that both receptors at the cell surface bind TNFa and TNFI3 with high affinity. The tissue distribution of the two receptors has not yet been systematically investigated, but it has become apparent that many cells and cell lines such as fibroblasts, or endothelial, myeloid and lymphoid cells, simultaneously express both receptors at various relative densities. A number of circumstantial findings support the view that the two receptors have distinct functional properties. First, the unrelated sequences of the intracellular regions of TNFRu and TNFR~3 in both man and mouse suggest different modes of signaling. However, signal transmission pathways which bypass the intracellular receptor sequences similar to the IL-6 receptor and gp 130 [44] cannot be ruled out. Second, a monoclonal antibody directed to the FAS antigen has agonistic functional activity and is able to mimic some, but not all types of cellular responses to TNF [45]. This might be due to an association of the FAS antigen with only one TNFR. Third, human TNF~t, in contrast to mouse TNFa, has a low systemic toxicity in mice, although murine cells in vitro bind and respond to human TNFct [46]. It has been shown that mouse TNFRa is not able to bind human TNF~t with high affinity [21]. Systemic toxicity of TNF in mice thus appears related to TNFR[~ rather than to TNFRa. Fourth, the strong induction of TNFRa, but
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
not of TNFR[3, seen early in activated human mononuclear cells and cell lines demonstrates the independent regulation of the TNFR and suggests distinct functions [ 13, 47]. However, there is clear evidence that the two TNFR in other instances may have equivalent functions. For example, in a reinvestigation of the immunomodulatory effects of TNF it was found that both TNFRa and TNFR[3 mediated the potent and late growthpromoting activity of TNFa in phytohemagglutinin-activated human mononuclear cells, even though TNFRa was expressed at about 10-fold higher level than TNFR[3 [48]. Interestingly, the full extent of cell activation depended on the activity of both receptors. Most of these studies concluded that both TNFR are functional. However, other investigators recently proposed that TNFR~3 is the only functional receptor for TNF. A monoclonal antibody, H398, which exclusively reacted with TNFR~3 was reported to completely block responses to TNF. This even occurred in cells in which TNFR[3 accounted for as little as about 15% of the total TNF binding sites [49]. Further studies will have to be carried out to reconcile these seemingly contradictory results. In summary, most of the presently available evidence suggests that depending on the cell type, or the functional state of the cell - the two TNFR may have similar or distinct functional properties.
Signal Transduction At present, little is known about the signal transmission pathways connected to TNFRa and TNFRJ3. TNF induces or activates several transcription factors. Both TNFR mediate the activation of the transcription factor NF-~cB [42, 47, 50]. TNF induces, for example, Jun, Fos, AP-1 and the interferon regulatory factors, IRF-1 and IRF-2 [51, 52]. The
pleiotropic nature of transcription factors and the multifactorial regulation of genes under the control of cytokines is illustrated by the regulation of IL-6 expression. IL-6 is induced by transcription factors binding to NF4cB-, AP-1, and NF-IL-6 response elements and is repressed by Fos [53-55]. Each of these transcription factors is in turn under the control of TNF. In addition, the 5'-flanking region of the IL-6 gene contains a further multiresponse element which confers direct inducibility by TNF [54]. The molecular cloning revealed that neither of the two TNFR contains sequences in the intracellular domains related to the catalytic domains of known protein kinases [reviewed in ref. 23]. It is therefore unlikely that they possess intrinsic protein kinase activity. However, several cellular proteins are phosphorylated in response to TNF, probably due to the activation of cellular kinases further down the transduction pathway [56-61]. The cellular responses to TNF have been reported to be mediated by two major signal transduction pathways, those activated by cAMP and by phospholipase C. TNF-activated protein kinase C appears to be involved in the induction of the jun and collagenase genes [5I]. Protein kinase A may be involved in the regulation of other genes [53], and the activation of other cellular kinases has been reported [60]. In other instances however, cellular responses to TNF such as the activation of the transcription factor NF4cB have been found to be independent of protein kinase C or cAMP [50, 62]. Furthermore, the phosphorylation involved in the transmodulation of the epidermal growth factor receptor has been found to be independent of protein kinase C [6l].
87
Concluding Remarks The identification and molecular cloning of the T N F R has advanced our understanding of the biological role of TNF. The relationship of the many different functions of TNF to one or both T N F R and the complexities of the TNF signal transduction pathway remain the
Qo
o ~ 1 7 6 1I o7 6 o l ,
o l o o ~
~
oo~
~176 ~
Qo
c o o p
9
o ~ 1 7 6 1 7 6
o e o l
o a o a ~ 1 7 6
, t a
foci of ongoing research. The TNF inhibitory proteins TNF-BP I and TNF-BP II have been recognized as truncated fragments of the respective receptors' extracellular domains; recombinant variants of TNF-BP may lead to new strategies in the treatment of human disease [40].
l o o p
oooo
o a ~
~
Q*
9 9 ~ 1 7 6 1 7 6 1 7 6~ 1 4 ~91 4 9
9 9
9
9
9
9
9 1 4 9 1 4 9
e9
References 1 Coley WB: The treatment of malignant tumors by repeated inoculations of erysipelas. With a report of ten original cases. Am J Med Sci 1893;105:487. 2 Carswell EA, Old L J, Kassel RL, Green S, Fiore N, Williamson B: An end 9 serum factor that causes necrosis of tumors. Proe Natl Acad Sci USA 1975;72:3666. 3 Old LJ: Tumor necrosis factor (TNF). Science 1985;230:630. 4 Beutler B, Cerami A: The biology of cachectin/TNF - a primary mediator of the host response. Ann Rev Immuno! 1989;7:625. 5 Tracey K, Fong Y, Hesse DG, Manogue KR, Lee AT, Kuo GC, Lowry SF, Cerami A: Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteriaemia. Nature 1987;330:662. 6 Mustafa MM, Ramilo O, Mertsola J, Risser RC, Beutter B, Hansen EJ, McCracken GH: Modulation of inflammation and cachectin activity in relation to treatment of experimental Hemophihts influenzae type b meningitis. J Infect Dis 1989;160: 818. 7 Nedwin GE, Naylor SL, Sakaguchi AY, Smith D, Jarett-Nedwin J, Pennica D, Goeddel DV, Gray PW: Human lymph 9 and tumor necrosis factor genes: Structure, homology and chromosomal location. Nucleic Acids Res 1985; 13:6361. 8 Eck M J, Sprang SR: The structure of tumor necrosis factor-alpha at 2.6 resolution. Implications for receptor binding. J Biol Chem 1989;264: 17595.
88
9 Jones EY, Stuart DI, Walker NPC: Structure of tumor necrosis factor. Nature 1989;338:225. 10 Schoenfeld H J, Poeschl B, Frey JR, Loetscher H, Hunziker W, Lustig A, Zulauf M: Efficient purification of recombinant human tumor necrosis factor beta from E. coli yields biologically active protein with a trimeric structure that binds to both TNF receptors. J Biol Chem 1991; 266:3863. 11 Loetscher HR, Schlaeger EJ, Lahm HW, Pan YCE, Lesslauer W, Brockhaus M: Purification and partial amino acid sequence analysis of two distinct tumor necrosis factor receptors from HL60 celts. J BioI Chem 1990;265:20131. 12 Loetscher HR, Pan YE, Lahm HW, Gentz R, Brockhaus M, Tabuchi H, Lesslauer W: Molecular cloning and expression of the 55kd tumor necrosis factor receptor. Cell 1990;61: 351. 13 Dembic Z, Loetscher HR, Gubler U, Pan YCE, Lahm HW, Gentz R, Brockhaus M, Lesslauer W: Two human TNF receptors have similar extracellular, but distinct intracellular, domain sequences. Cytokine 1990; 2:231. 14 SchaU TJ, Lewis M, Koller KJ, Lee A, Rice GC, Wong GHW. Gatanaga T, Granger GA, Lentz R, Raab H, Kohr W J, Goeddel DV: Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell 1990;61:361.
15 Smith CA, Davis T, Anderson D, Solam L, Beckmann MP, Jer-zy R, Dower SK, Cosman D, Goodwin RG: A receptor for tumor necrosis factor defines an unusual family of cellular and viral proteins. Science 1990;248:1019. 16 Heller RA, Song K, Onasch MA, Fischer WH, Chang D, Ringold GM: Complementary DNA cloning of a receptor for tumor necrosis factor and demonstration of a shed form of the receptor. Proc Natl Acad Sci USA 1990;87:6151. 17 Kohno T, Brewer MT, Baker SL, Schwartz PE, King MW, Hale KK, Squires CH, Thompson RC, Vannice JL: A second tumor necrosis factor receptor gene product can shed a naturally occurring tumor necrosis factor inhibitor. Proc Natl Acad Sci USA 1990:87:8331. 18 Nophar Y, Kemper O, Brakebusch C, Engelmann H. Zwang R, Aderka D, Holtmann H, Wallach D: Soluble forms of tumor necrosis factor receptors (TNF-Rs). The cDNA for the type I TNF-R cloned using amino acid sequence data of its soluble form encodes both the cell surface and a soluble form of the receptor. EMBO J 1990;9:3269. 19 Himmler A, Maurerfogy I, Kronke M, Scheurich P, Pfizenmaier K, Lantz M, Olsson I, Hauptmann C, Stratowa C, Adolf GR: Molecular cloning and expression of human and rat tumor necrosis factor receptor chain (p60) and its soluble derivative - tumor necrosis factor binding protein DNA. Cell Biol 1990;9: 705.
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
20 Gray PW, Barrett K, Chantry D, Turner M, Feldmann M: Cloning of human tumor necrosis factor cDNA and expression of recombinant soluble TNF-binding protein. Proc Natl Acad Sci USA 1990;87:7380. 21 Lewis M, Tartaglia LA, Lee A, Bennett GL, Rice GC, Wong GHW, Chen EY, Goeddel DV: Cloning and expression of cDNAs for two distinct murine tumor necrosis factor receptors demonstrate one receptor is species specific. Proc Natl Acad Sci USA 1991;88:2830_ 22 Rothe JG, Brockhaus M, Gentz R, Lesslauer W: Molecular cloning and expression of the mouse TNF receptor type b. Immunogenelics, in press. 23 Loetscher HR, Brockhaus M, Dembic Z, Gentz R, Gubler U, Hohmann HP, Lahm H W van Loon APGM, Pan YCE, Schlaeger EJ, Steinmetz M, Tabuchi H, Lesslauer W: Two distinct tumor necrosis factor receptors - members of a new cylokine receptor gone family; in MacIean (ed): Oxford Survey on Eucaryotic Genes. Oxford, Oxford University Press. 1991. 24 Hohmann H, Remy R, 8rockhaus M, van Loon APGM: Two different cell types have different major receptors for human tumor necrosis factor (TNF alpha). I Biol Chem 1989;264:14927. 25 Brockhaus M, Schoenfeld H J, Schlaeger EJ, Hunziker W, Lesslauer W, Loetscher HR: Identification of two types of tumor necrosis factor receptors on human cell lines by monoclonal antibodies. Proc Natl Acad Sci USA 1990;87:3127. 26 Loetscher HR, Schlaeger EJ, Lahm HW, Pan YCE, Lesslauer W, Brockhaus M: Purification and partial amino acid sequence analysis of two distinct tumor necrosis factor receptors from HL60 cells. J Biol Chem 1990;265:20131, 27 Loetseher H: Purification and structural properties of two distinct TNF receptors on h u m a n cells; in Aggarwal, Vilcek (eds): Tumor necrosis factors - structure, function and mechanism of action; Marcel Dekker, New York, ! 991.
28 Mallett S, Barclay AN: A new superfamily of cell surface proteins related to the nerve growth factor receptor, lmmunol Today 1991;12: 220. 29 Rogers S, Wells R, Rechsteiner M: Amino acid sequences common to rapidly degraded proteins. The PEST hypothesis. Science 1986;234: 364. 30 Seckinger P, Isaaz S, Dayer JM: A human inhibitor of tumor necrosis factor alpha. J Exp Med 1988;167: 1511. 31 Seckinger P, Isaaz S, Dayer JM: Purification and biologic characterisation of a specific tumor necrosis factor alpha inhibitor. J Biol Chem 1989;264: t 1966. 32 Olsson I, Lantz M, Nilsson E, Peetre C, Thysell H, Grubb A, Adolf G: Isolation and characterisation of a tumor necrosis factor binding protein from urine. Eur J Haematol 1989;42:270. 33 Peetre C, Thyseel H, Grubb A, Olsson I: A tumor necrosis factor binding protein is present in human biological Huids. Eur J Haematol 1988; 41;414. 34 Engelmann H, Aderka D, Rubinstein M, Rotman D, Wallach D: A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem 1989;264:11974. 35 Engelmann H, Novick D, Wallach D: 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 1990;265:1531. 36 Olsson I, Lantz M, Nilsson E, Peetre C, ThyseU H, Grubb A, Adolf G: Isolation and characterisation of a tumor necrosis factor binding protein from urine. Eur J Haematol 1989;42:270. 37 Seckinger P, Vey E, Turcatti G, Wingfield P, Dayer JM: Tumor necrosis factor inhibitor. Purification, NH2-terminal amino acid sequence and evidence for anti-inflammatory and immunomodulatory activities. EurJ Immuno11990;20:1167.
38 Digel W, Por'zsolt F, Lesslauer W, Brockhaus M: Evidence for soluble receptors for tumor necrosis factor in the serum of cancer patients high levels in B-cell malignancies, submitted. 39 Gatanaga T, Hwang C, Kohr W, Capuccini F, Lucci IA, Jeffes EWB, Lentz R, Tomich J, Yamamoto RS, Granger GA: Purification and characterisation of an inhibitor (soluble tumor necrosis factor receptor) for tumor necrosis factor and lymphotoxin obtained from the serum ultrafiltrates of human cancer patients. Proc Natl Acad Scl USA 1990;87: 8781. 40 Lesslauer W, Tabuchi H, Gentz R, Schlaeger EJ, Brockhaus M, Grau G, Piguet PF, Pointaire P, Vassalli P, Loetscher H: Bioactivity of recombinant human TNF receptor fragments. J Cell Biochem 1991; 15F(suppl): I 15. 41 Loetscher HR, Gentz R, Zulauf M, Lustig A, Tabuchi H. Schlaeger E J. Brnekhaus M, Gallati H, Manneberg M, Lesslauer W: Recombinant 55kDa TNF receptor. Stoichiometry of binding to TNFalpha and TNF'beta and inhibition o f T N F activity~ J Biol Chem, in press. 42 Hohmann HP, Remy R, Poeschl B, van Loon APGM: Tumor necrosis factors-alpha and -beta bind to the same two types of tumor necrosis factor receptors and maximally activate the transcription factor NK-kB at low receptor occupancy and within minutes after receptor binding. J Biol Chem 1990;265:15183. 43 Engelmann H, Holtmann H, Brakebusch C, Avni SY, Sarov I, Nophar Y, Hadas E, Leitner O, Wallach D: Antibodies to a soluble form of a tumor necrosis factor (TNF) receptor have TNF-like activity. J BiN Chem 1990;265:14497. 44 Hibi M, Murakami M, Saito M, Hirano T, Taga T, Kishimoto T: Molecular cloning and expression of an IL-6 signal transducer, gpl30. Cell 1990;63:1149. 45 Kobayashi N, Hamamoto Y, Yamamoto N, Ishii A, Yonehara M, Yonehara S: Anli-FAS monoclonal antibody is cytocidal to human immunodeficiency virus-infected ceiis without augmenting viral replication. Proc Natl Acad Sci USA 1990; 87:9620.
89
46 Brouckaert PG, Everaerdt B, Libert C, Takahashi N, Fiers W: Species specificity and involvement of other cytokines in endotoxic shock action of recombinant tumor necrosis factor in mice. Agents Actions 1989;26: 196. 47 Hohmann HP, Brockhaus M, Baeuerle PA, Remy R, Kolbeck R, van Loon APGM: Expression of the types A and B tumor necrosis factor (TNF) receptors is independently regulated and both receptors mediate activation of the transcription factor NF-kB. TNFalpha is not needed for induction of a biological effect via TNF receptors. J Biol Chem 1990;265:22409. 48 Gehr G, Gentz R, Loetscher HR, Brockhaus M, Tabuchi H, Lesslauer W: TNF receptors alpha and beta (TNFRalpha, TNFRbeta) in human mononuclear cell activation, submitted. 49 Thoma B. Grell M, Pfizenmaier K, Scheurich P: Identification of a 60 kD tumor necrosis factor (TNF) receptor as the major signal transducing component in TNF responses. J Exp Med 1990;172:1019. 50 Meichle A, Schuetze S, Hensel G, Brunsing D, Kroenke M: Protein kinase C-independent activation of nuclear factor kB by tumor necrosis factor. J Biol Chem 1990;265:8339.
51 Brenner DA, O'Hara M, Angel P, Chojkier M, Karin M: Prolonged activation ofjun and collagenase genes by tumor necrosis factor alpha. Nature 1989:337:661. 52 Harada H, Fujita T, Miyamoto M, Kimura Y, Maruyama M, Furia A, Miyata T, Taniguchi T: Structurally similar but functionally distinct factors IRF-1 and IRF-2 bind to the same regulatory elements of IFN and IFN-inducible genes. Cell 1989; 58:729. 53 Zhang Y, Lin JX, Vilcek J: Interleukin 6 induction by tumor necrosis factor and interleukin 1 in human fibroblasts involves activation of a nuclear factor binding to a kB like sequence. Mol Cell Biol 1990;10: 3818. 54 Ray A, Sassone-Corsi P, Sehgal P: A multiple cytokine- and second messenger-responsive element in the enhancer of the human interleukin-6 gene. Similarities with c-fos gene regulation. Mol Cell Biol 1989;9: 5537. 55 Akira S, Isshiki H, Sugita T, Tanabe O, Kinoshita S, Nishio Y, Nakajima T, Hirano T, Kishimoto T: A nuclear factor for IL-6 expression (NFIL6) is a member ofa C/EBP family. EMBO J 1990;9:1897. 56 Schuetze S, Scheurich P, Pfizenmaier K, Kroenke M: Tumor necrosis factor signal transduction. Tumor specific serine phosphorylation of a 26 kDa cytosolic protein. J Biol Chem 1989;264:3562.
57 Donato NJ, Gallick GE, Steck PA, Rosenblum MG: Tumor necrosis factor modulates epidermal growth factor receptor phosphorylation and kinase activity in human tumor cells. Correlation with cytotoxicity. J Biol Chem 1989;264:20474. 58 Marino M, Pfeffer LM, Guidon PT, Donner DB: Tumor necrosis factor induces phosphorylation of a 28 kDa mRNA cap-binding protein in human cervical carcinoma cells. Proc Natl Acad Sci USA 1989;86: 8417. 59 Kohno M, Nishizawa N, Tsujimoto M, Nomoto H: Mitogenic signalling pathway of tumor necrosis factor involves the rapid tyrosine phosphorylation of 41'000 M~ and 43"000 Mr cytosol proteins. Biochem J 1990; 267:91. 60 Shiroo M, Matsushima K: Enhanced phosphorylation of 65 and 74 kDa proteins by tumor necrosis factor and interleukin 1 in human peripheral blood mononuclear cells. Cytokine 1990;2:13. 61 Bird TA, Saklatvala J: IL-I and TNF transmodulate epidermal growth factor receptors by a protein kinase C-independent mechanism. J lmmunol 1989;142:126. 62 Hohmann HP, Remy R, Schneidereit C, van Loon APGM: Cyclic AMP independent activation of the transcription factor NF-kB in HL60 cells by tumor necrosis factors alpha and beta. Mol Cell BioI [991;11: 2315.
90
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
Joachim Rothe G&ela Gehr Hansruedi Loetscher Werner Lesslauer
Tumor NecrosisFactor ReceptorsStructure and Function
F. Hoffmann-LaRoche, Ltd., Basel, Switzerland
oooo*loo~176176176176176176
tooo*eeoooel.*..~176176176176176176
Key Words
Abstract
Tumor necrosis factor, receptors and inhibitory proteins Nerve growth factor receptor family Cytokines
Tumor necrosis factors (TNFs) have been a focus of research for well over a decade now. The identification and recent molecular cloning of two different types of ceil-surface TNF receptors will shed further light on the mode of action of these pleiotropic cytokines. In the present article, we summarize the data on the biochemistry and structure of the receptors and focus on the molecular cloning of the respective cDNAs. The nucleotide sequences of the receptor genes revealed that both TNF receptors belong to the still growing nerve growth factor receptor gene family. The function and origin of TNF inhibitory proteins as well as receptor-mediated signal transduction are discussed.
JeloooooooooQo*ooo~176176176176176176
* e e e e e e e o * e e e e e e e
The functional activities of cells in immunological and inflammatory host defense reactions are mediated by a large and still growing number of protein factors such as interferons, interleukins, colony-stimulating factors and tumor necrosis factors (TNF), which are commonly referred to as cytokines. Cytokines act by binding to specific cell surface receptors. Their mode of action in general is local and paracrine rather than sys-
temic, and their serum concentrations usually are very low. Typically, a cytokine is produced by several different cell types and its receptors are expressed on a variety of potential target cells. It may elicit the production of secondary cytokines in a cascade-like fashion or modulate cytokine receptor function. Specific mechanisms such as the controlled release of cytokine inhibitors limit cytokine activities. A complex regulatory network of me-
Dr. W. Lesslauer Building 15, Room 6 F. Hoffmann-La Roche Ltd. -CH-4002 Basel (Switzerland)
9 1992 S. Karger AG, Basel 0257-277X/92/ 0112-0081 $2.75/0
diators thus is generated which forms the basis of the pleiotropic and redundant cytokine functions. TNF was first discovered by its ability to induce hemorrhagic necrosis of certain tumors [ l, 2]. Later, after the molecular cloning and recombinant expression, a wide range of biological activities was discovered. TNF is now considered to be one of the most multifunctional response modifiers in inflammatory processes and immunological reactions, to have growth factor activity on some cell types and to induce expression of adhesiontype or histocompatibility molecules on others [for reviews, see ref. 3, 4]. In some instances, extremely harmful effects are produced by TNF. For example, the lethal consequences of septic shock are largely mediated by TNF [5]. Furthermore, tissue damage in bacterial meningitis correlates with TNF concentration in cerebrospinal fluid [6]. The term TNF commonly refers to the two related factors, TNFa (cachectin) and TNF~ (lymphotoxin), which have a sequence homology of about 30% [7]. TNFa and TNF[3 are thought to be mainly produced by monocytes/macrophages and activated T lymphocytes. The three-dimensional structure of TNFa has been investigated in X-ray diffraction studies [8, 9]. TNFa crystallizes as a trimer, and the biologically active forms of both TNFa and TNFI3 are believed to be trimers [ 10]. Recently, several groups identified two distinct TNF receptors (TNFR), termed TNFRa (TNFR type II, p75, utr antigen) and TNFR[3 (TNFR type I, p55, htr antigen) in man [1120] and mouse [21, 22]. When the predicted amino acid sequences of the two receptors were analyzed, a high degree of similarity of the extracellular regions emerged, whereas the sequences of the intracellular regions were entirely unrelated. These findings raised the questions of whether the two receptors have distinct functional properties, and how the
82
many different activities of TNF relate to TNFRct and TNFRI3. The present review focuses on the structure of the two receptors and summarises current views of their biological activities.
Protein Biochemistry of TNFR In the past few years TNFR have been identified on the surface of most human and murine cells analysed in a large number of studies [for references, see ref. 23]. In radioligand binding studies, typically a single class of cellular TNF binding sites with Ka values in the picomolar range was found. Chemical crosslinking of radioiodinated TNF to cell surface binding sites led to the identification of various TNFa- and TNFI3-reactive bands in SDS-PAGE from about 50 to 138 kD. The analysis of the crosslinked complexes from various cells indicated at least two distinct TNFR molecules which differ not only in molecular mass but also in ligand-binding affinity, glycosylation, immunoreactivity and proteolytic fingerprints [24]. TNFRa has a molecular mass of 75 kD, TNFR[3 is smaller (55 kD). The existence of two different receptors was confirmed by monoclonal antibodies raised against partially purified TNFR fractions; two classes of antibodies were found which reacted exclusively with either TNFRa or TNFR[3 [25]. The receptors are expressed in different relative amounts on various cell types in the range of several hundred to a few thousand molecules per cell. The two TNFR were purified from the HL60 and U937 cell lines, and from human placenta by combined immuno- and ligandaffinity column chromatography and reversephase high performance liquid chromatography [23, 26, 27]. Both receptors were found to bind TNFa and TNF[3 with Ka values in the picomolar range, indicating that no accessory
Rothe/Gehr/Loetscher/Les slauer
TNF Receptors
protein is necessary to form a high-affinity ligand-binding site with either TNFR [10]. The receptor proteins purified from HL60 cells revealed major bands at 75 kD for TNFRa and 55 kD for TNFRI] in addition to a few minor bands (see below) by SDS-PAGE analysis. Specific binding of 12SI-TNFa to all these bands was demonstrated in ligand blot experiments [11]. Furthermore, the 75- and 55-kD bands reacted exclusively with monoclonal antibodies specific for TNFRa and TNFRI3, respectively [11, 25]. Partial N-terminal and internal peptide sequences were determined for the 75kD and 55kD bands. Protein sequencing of the minor bands which copurified with TNFRct and TNFR~ revealed that they were derivatives or fragments of the receptors [13, 26]. Interestingly, a somewhat shortened form of TNFRa was found to be ubiquitinated; the functional significance of the ubiquitination remains to be established.
Molecular Cloning of the Two TNFR
Several groups have reported the molecular cloning of human [12-20] and mouse TNFRct and TNFRJ3 [21, 22]. In our own work we have isolated the cDNAs encoding human TNFRa and TNFR[3 using partial amino acid sequence information and a combination of polymerase chain reaction and conventional cloning strategies [12, 13]. The mouse TNFR cDNAs were isolated from murine cDNA libraries by interspecies crosshybridization using a fragment of the homologous human cDNA as probes [21, 22]. The open reading frame of the human TNFRJ3 cDNA encodes a protein of 455 amino acids with a typical signal peptide of 29 amino acids and a single 21-amino acid transmembrane region, which separates the extracellular Nterminal domain (182 residues including 24 cysteines) from the C-terminal cytoplasmic
domain (223 residues). Three potential Nlinked glycosylation sites are contained in the extracellular domain. The amino terminal region of the molecule is extracellularly located, because the recombinant N-terminal region binds TNFet and antibodies which bind to intact cells. Leucine was determined to be the N-terminal residue by protein sequencing. Mature human TNFRI3 therefore consists of 426 amino acids with a theoretical molecular mass of 47.5 kD. The human TNFR~ cDNA encodes a receptor protein of 439 amino acids with a single transmembrane region and a theoretical molecular mass of nearly 50 kD. The extracellular and intracellular regions are composed of 235 and 178 residues, respectively. Two potential N-linked glycosylation sites are present in the extracellular domain. Posttranslational modifications of both receptors account for the differences between the theoretical and the apparent molecular weights estimated from SDS-PAGE. A hypothetical overall structure of both receptor molecules is schematically illustrated in figure I. Transfection of either of the two human TNFR cDNAs into COS-1 cells led to the transient expression of a single class of highaffinity TNFa binding sites. Apparent Kd were determined with the transfected COS-1 cells by radioligand binding assays; values of about 0.1 nM(TNFRa) and 0.5 nM(TNFRI3) were found [ 12, 13]. The predicted amino acid sequences of the two mouse TNFR reveal high degrees of similarity to the homologous human receptors with regard to overall size, relative size of domains and potential glycosylation pattern. The high degree of sequence conservation between the human and mouse TNFR is most clearly demonstrated by the fact that, as shown in table 1, the amino acid sequence identity in the extracellular domains of human and mouse TNFRct and TNFR[3 reaches 58 and 70 %, respectively [ 12-15, 21, 22].
83
TNFR Belong to a New Receptor Gene Family
ct M,H
i I
ii II H M,H
iii fll
H M,H
iv
IV
Cytoplasm
Fig. 1. Schematic representation of TNFRa and TNFR!3. Cysteine-riche repeats are designated by Roman numbers, conserved cysteine residues itself by horizontal lines inside the repeats. Circles are potential N-linked glycosylation sites in man (H) and mouse (M).
Table 1. Percent amino acid identity between corresponding parts of TNF receptors in man (h) and mouse (m)
Leader EC domain TM region IC domain
hTNFRttmTNFRct
hTNFRI]mTNFRI3
TNFRctTNFRI3
55 58 65 73
76 70 74 59
19 22 28 < 10
Values given in the last section are approximate numbers. EC = Extracellular, TM = transmembrane and IC = intracellular,
84
The predicted amino acid sequences of the extracellular regions of the two TNFR in man and mouse reveal significant similarities. The most prominent feature is a distinct pattern of cysteine residues defining four conserved sequence repeats of roughly forty amino acids shown schematically in figure t. The two N-terminal repeats of both receptors contain six cysteines with a consensus pattern ofCX]0_~ sCX0_2CXaCXs_1 ICX3-8C. The third and fourth repeats contain six (TNFR[~) or four cysteines (TNFRa) in homologous positions and thus are less well conserved between the receptors. The alignment, however, still scores significantly above a random score with the mutation data matrix. Five other currently known sequences have striking similarities with the TNFR extracellular regions: the human and rat low-affinity nerve growth factor receptors, the B cell surface antigen CD40, the murine cDNA clone 4-1 BB isolated from induced helper and cytolytic T cell clones, the transcriptionally open reading frame T2 from the Shope fibroma virus, and the OX 40 antigen present on activated CD4-positive rat Tlymphocytes [for references, see ref. 28]. The similarity among all these receptor-type molecules, however, is restricted to the presumably heavily disulfide-bonded extracellular regions. Interestingly, the level of sequence identity between the two TNFR is lower than that to other members of the receptor gene family. The evolutionary conservation of the cysteine-rich repeat motif suggests that this sequence folds into a three-dimensional scaffold, which is favorable for the binding of protein ligands. The fine specificity for binding a specific ligand according to this hypothesis is determined by the intervening amino acids. Electrostatic interactions may play a role. It is noteworthy that both TNFR carry an overall
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
positive charge at physiologic pH, while TNF is negatively charged. In contrast, the negatively charged extracellular domain of the NGFR is complementary to the positively charged NGF. The intracellular regions of the two TNFR do not show any obvious sequence similarity to each other nor to any other known mammalian sequence. In particular, no similarities to catalytic domains of tyrosine- or serine/ threonine-specific protein kinases are present. There are some patterns which resemble sequence elements involved in signal transduction in other molecules, i.e. potential phosphorylation sites for tyrosine kinase (TNFRI3), protein kinase C (TNFRct and TNFRI3) and cyclic nucleotide dependent kinase (TNFRct and TNFR[3). Interestingly, the cytoplasmic domains of both TNF receptors have a high content in serine, threonine and proline. These so-called PEST sequences might relate to high turnover rates and intracellular protein degradation [29].
TNF Inhibitory Proteins A few years ago, two proteins were discovered in human serum and urine which inhibited TNF activity in cell culture. They were termed TNF-BP I and II and shown to act by binding to TNFct and, at least for TNF-BP I, to a lesser extent to TNFI3 [30-35]. When partial amino acid sequences of the TNF-BP had been determined, both TNF-BP were recognized as truncated fragments of the extracellular regions of the respective TNFR [35-37]. The serum concentrations of the TNF-BP have not yet been systematically investigated, but in one report they have been found to occur at concentrations in the range of ng/ml [38]. There is general agreement that normal TNFct concentrations in serum are about 30 pg/ml or lower. TNF-BP are therefore present
in large molar excess, but their physiologic role is not fully understood. They may act by neutralization of TNF in the systemic circulation, thus enforcing paracrine type activity. Alternatively, they may be viewed as a buffer from which TNF is slowly released. The view that they exert a TNF-neutralizing activity is supported by animal studies in which natural TNF-BP [39] or recombinant TNFR molecules [40] have been found to inhibit TNFmediated pathology.
Recombinant Soluble TNFR The extracellular regions of both TNFR have been expressed in recombinant soluble form (rec sol TNFR) in the baculovirus/Sf9 or CHO cell expression systems and the TNF binding properties were determined [41]. With rec sol TNFRI3, a single class of highaffinity binding sites for TNFct with K,~ values between 0.3 and 0.4 ru'v/was found. In contrast to the native cell surface-bound receptor which does not distinguish between TNFct and TNF[3 [24, 42], rec sol TNFRI3 showed a significantly lower affinity for TNFI3 compared to that for TNFcc The Kd values ranged from 0.5 to 1.6 nM depending on the expression system used. This is consistent with the previously reported finding that the naturally occurring soluble TNF-BP I is less potent in neutralizing TNFI3 [31, 35]. Interestingly, the Scatchard analysis of TNFct binding to rec sol TNFRct in a solidphase binding assay revealed a nonlinear curve which suggested two affinity classes for ligand binding [Loetscher et al., unpubl, data]. The best fit to the binding curve was achieved with Kd of 0.2 nM (high affinity site) and 2.5 nM (low affinity site). The analogous Kd values for the binding of TNFfl to rec sol TNFRct were in the range of 0.2-0.6 and 3.7-10 nM for the high and low affinity sites, respective-
85
ly. The significance of the low-affinity TNF binding site of rec sol TNFRa remains to be established. It is noteworthy that a single class of binding sites of membrane-bound TNFRct has been reported in a number of studies. It might be proposed that the low-affinity site is generated by a conformational change induced by the free C-terminal end of rec sol TNFRa. However, both high and low affinity TNFa binding sites with Kd values similar to those described above were recently reported for both native and recombinant full-length TNFRct [15]. Chimeric proteins in which the extracellular domains of human TNFRa and TNFRI3 were fused to the hinge region of human IgG y3 (rsTNFR~t-h73 and rsTNFRI3-hy3) have been expressed in mouse myeloma cells [41]. These fusion proteins have a disulfidebonded dimeric structure similar to that of an antibody molecule. In ligand-binding studies it was found that the fusion proteins bind TNFct and TNFI3 with high affinity. A linear relationship was obtained in the Scatchard analysis with Ka values between 0.1 and 0.2 n_M for both TNFct and TNFI3. The various rec sol TNFR constructs were found to compete with full-length TNFRct and TNFR[3 for TNF binding in in vitro binding assays. Furthermore, it could be shown that the cellular cytotoxicity of TNF in cell culture is inhibited by rec sol TNFR in a concentration-dependent fashion. The comparison of the TNFneutralizing activity demonstrated that rsTNFRf3-h73 inhibits TNF bioactivity at 10to 100-fold lower concentrations than rsTNFRI3. The molecular mass of rsTNFRI3 in solution as determined by light scattering and analytical ultracentrifugation studies is about 25 kD [41]. Stable complexes of rsTNFRI3 with TNFa and TNFI3 were formed and were both found to have a molecular mass of about 140 kD, suggesting a stoichiometry of three
86
rsTNFRI3 bound to one TNF trimer. The stability of these complexes is of interest since microclustering of TNFRI3 has been proposed as an essential step in TNF signal generation at the surface of target cells [43].
Functional Aspects The discovery of two distinct TNFR raised the question whether they have equivalent or distinct functions. It was soon realized, that both receptors at the cell surface bind TNFa and TNFI3 with high affinity. The tissue distribution of the two receptors has not yet been systematically investigated, but it has become apparent that many cells and cell lines such as fibroblasts, or endothelial, myeloid and lymphoid cells, simultaneously express both receptors at various relative densities. A number of circumstantial findings support the view that the two receptors have distinct functional properties. First, the unrelated sequences of the intracellular regions of TNFRu and TNFR~3 in both man and mouse suggest different modes of signaling. However, signal transmission pathways which bypass the intracellular receptor sequences similar to the IL-6 receptor and gp 130 [44] cannot be ruled out. Second, a monoclonal antibody directed to the FAS antigen has agonistic functional activity and is able to mimic some, but not all types of cellular responses to TNF [45]. This might be due to an association of the FAS antigen with only one TNFR. Third, human TNF~t, in contrast to mouse TNFa, has a low systemic toxicity in mice, although murine cells in vitro bind and respond to human TNFct [46]. It has been shown that mouse TNFRa is not able to bind human TNF~t with high affinity [21]. Systemic toxicity of TNF in mice thus appears related to TNFR[~ rather than to TNFRa. Fourth, the strong induction of TNFRa, but
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
not of TNFR[3, seen early in activated human mononuclear cells and cell lines demonstrates the independent regulation of the TNFR and suggests distinct functions [ 13, 47]. However, there is clear evidence that the two TNFR in other instances may have equivalent functions. For example, in a reinvestigation of the immunomodulatory effects of TNF it was found that both TNFRa and TNFR[3 mediated the potent and late growthpromoting activity of TNFa in phytohemagglutinin-activated human mononuclear cells, even though TNFRa was expressed at about 10-fold higher level than TNFR[3 [48]. Interestingly, the full extent of cell activation depended on the activity of both receptors. Most of these studies concluded that both TNFR are functional. However, other investigators recently proposed that TNFR~3 is the only functional receptor for TNF. A monoclonal antibody, H398, which exclusively reacted with TNFR~3 was reported to completely block responses to TNF. This even occurred in cells in which TNFR[3 accounted for as little as about 15% of the total TNF binding sites [49]. Further studies will have to be carried out to reconcile these seemingly contradictory results. In summary, most of the presently available evidence suggests that depending on the cell type, or the functional state of the cell - the two TNFR may have similar or distinct functional properties.
Signal Transduction At present, little is known about the signal transmission pathways connected to TNFRa and TNFRJ3. TNF induces or activates several transcription factors. Both TNFR mediate the activation of the transcription factor NF-~cB [42, 47, 50]. TNF induces, for example, Jun, Fos, AP-1 and the interferon regulatory factors, IRF-1 and IRF-2 [51, 52]. The
pleiotropic nature of transcription factors and the multifactorial regulation of genes under the control of cytokines is illustrated by the regulation of IL-6 expression. IL-6 is induced by transcription factors binding to NF4cB-, AP-1, and NF-IL-6 response elements and is repressed by Fos [53-55]. Each of these transcription factors is in turn under the control of TNF. In addition, the 5'-flanking region of the IL-6 gene contains a further multiresponse element which confers direct inducibility by TNF [54]. The molecular cloning revealed that neither of the two TNFR contains sequences in the intracellular domains related to the catalytic domains of known protein kinases [reviewed in ref. 23]. It is therefore unlikely that they possess intrinsic protein kinase activity. However, several cellular proteins are phosphorylated in response to TNF, probably due to the activation of cellular kinases further down the transduction pathway [56-61]. The cellular responses to TNF have been reported to be mediated by two major signal transduction pathways, those activated by cAMP and by phospholipase C. TNF-activated protein kinase C appears to be involved in the induction of the jun and collagenase genes [5I]. Protein kinase A may be involved in the regulation of other genes [53], and the activation of other cellular kinases has been reported [60]. In other instances however, cellular responses to TNF such as the activation of the transcription factor NF4cB have been found to be independent of protein kinase C or cAMP [50, 62]. Furthermore, the phosphorylation involved in the transmodulation of the epidermal growth factor receptor has been found to be independent of protein kinase C [6l].
87
Concluding Remarks The identification and molecular cloning of the T N F R has advanced our understanding of the biological role of TNF. The relationship of the many different functions of TNF to one or both T N F R and the complexities of the TNF signal transduction pathway remain the
Qo
o ~ 1 7 6 1I o7 6 o l ,
o l o o ~
~
oo~
~176 ~
Qo
c o o p
9
o ~ 1 7 6 1 7 6
o e o l
o a o a ~ 1 7 6
, t a
foci of ongoing research. The TNF inhibitory proteins TNF-BP I and TNF-BP II have been recognized as truncated fragments of the respective receptors' extracellular domains; recombinant variants of TNF-BP may lead to new strategies in the treatment of human disease [40].
l o o p
oooo
o a ~
~
Q*
9 9 ~ 1 7 6 1 7 6 1 7 6~ 1 4 ~91 4 9
9 9
9
9
9
9
9 1 4 9 1 4 9
e9
References 1 Coley WB: The treatment of malignant tumors by repeated inoculations of erysipelas. With a report of ten original cases. Am J Med Sci 1893;105:487. 2 Carswell EA, Old L J, Kassel RL, Green S, Fiore N, Williamson B: An end 9 serum factor that causes necrosis of tumors. Proe Natl Acad Sci USA 1975;72:3666. 3 Old LJ: Tumor necrosis factor (TNF). Science 1985;230:630. 4 Beutler B, Cerami A: The biology of cachectin/TNF - a primary mediator of the host response. Ann Rev Immuno! 1989;7:625. 5 Tracey K, Fong Y, Hesse DG, Manogue KR, Lee AT, Kuo GC, Lowry SF, Cerami A: Anti-cachectin/TNF monoclonal antibodies prevent septic shock during lethal bacteriaemia. Nature 1987;330:662. 6 Mustafa MM, Ramilo O, Mertsola J, Risser RC, Beutter B, Hansen EJ, McCracken GH: Modulation of inflammation and cachectin activity in relation to treatment of experimental Hemophihts influenzae type b meningitis. J Infect Dis 1989;160: 818. 7 Nedwin GE, Naylor SL, Sakaguchi AY, Smith D, Jarett-Nedwin J, Pennica D, Goeddel DV, Gray PW: Human lymph 9 and tumor necrosis factor genes: Structure, homology and chromosomal location. Nucleic Acids Res 1985; 13:6361. 8 Eck M J, Sprang SR: The structure of tumor necrosis factor-alpha at 2.6 resolution. Implications for receptor binding. J Biol Chem 1989;264: 17595.
88
9 Jones EY, Stuart DI, Walker NPC: Structure of tumor necrosis factor. Nature 1989;338:225. 10 Schoenfeld H J, Poeschl B, Frey JR, Loetscher H, Hunziker W, Lustig A, Zulauf M: Efficient purification of recombinant human tumor necrosis factor beta from E. coli yields biologically active protein with a trimeric structure that binds to both TNF receptors. J Biol Chem 1991; 266:3863. 11 Loetscher HR, Schlaeger EJ, Lahm HW, Pan YCE, Lesslauer W, Brockhaus M: Purification and partial amino acid sequence analysis of two distinct tumor necrosis factor receptors from HL60 celts. J BioI Chem 1990;265:20131. 12 Loetscher HR, Pan YE, Lahm HW, Gentz R, Brockhaus M, Tabuchi H, Lesslauer W: Molecular cloning and expression of the 55kd tumor necrosis factor receptor. Cell 1990;61: 351. 13 Dembic Z, Loetscher HR, Gubler U, Pan YCE, Lahm HW, Gentz R, Brockhaus M, Lesslauer W: Two human TNF receptors have similar extracellular, but distinct intracellular, domain sequences. Cytokine 1990; 2:231. 14 SchaU TJ, Lewis M, Koller KJ, Lee A, Rice GC, Wong GHW. Gatanaga T, Granger GA, Lentz R, Raab H, Kohr W J, Goeddel DV: Molecular cloning and expression of a receptor for human tumor necrosis factor. Cell 1990;61:361.
15 Smith CA, Davis T, Anderson D, Solam L, Beckmann MP, Jer-zy R, Dower SK, Cosman D, Goodwin RG: A receptor for tumor necrosis factor defines an unusual family of cellular and viral proteins. Science 1990;248:1019. 16 Heller RA, Song K, Onasch MA, Fischer WH, Chang D, Ringold GM: Complementary DNA cloning of a receptor for tumor necrosis factor and demonstration of a shed form of the receptor. Proc Natl Acad Sci USA 1990;87:6151. 17 Kohno T, Brewer MT, Baker SL, Schwartz PE, King MW, Hale KK, Squires CH, Thompson RC, Vannice JL: A second tumor necrosis factor receptor gene product can shed a naturally occurring tumor necrosis factor inhibitor. Proc Natl Acad Sci USA 1990:87:8331. 18 Nophar Y, Kemper O, Brakebusch C, Engelmann H. Zwang R, Aderka D, Holtmann H, Wallach D: Soluble forms of tumor necrosis factor receptors (TNF-Rs). The cDNA for the type I TNF-R cloned using amino acid sequence data of its soluble form encodes both the cell surface and a soluble form of the receptor. EMBO J 1990;9:3269. 19 Himmler A, Maurerfogy I, Kronke M, Scheurich P, Pfizenmaier K, Lantz M, Olsson I, Hauptmann C, Stratowa C, Adolf GR: Molecular cloning and expression of human and rat tumor necrosis factor receptor chain (p60) and its soluble derivative - tumor necrosis factor binding protein DNA. Cell Biol 1990;9: 705.
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
20 Gray PW, Barrett K, Chantry D, Turner M, Feldmann M: Cloning of human tumor necrosis factor cDNA and expression of recombinant soluble TNF-binding protein. Proc Natl Acad Sci USA 1990;87:7380. 21 Lewis M, Tartaglia LA, Lee A, Bennett GL, Rice GC, Wong GHW, Chen EY, Goeddel DV: Cloning and expression of cDNAs for two distinct murine tumor necrosis factor receptors demonstrate one receptor is species specific. Proc Natl Acad Sci USA 1991;88:2830_ 22 Rothe JG, Brockhaus M, Gentz R, Lesslauer W: Molecular cloning and expression of the mouse TNF receptor type b. Immunogenelics, in press. 23 Loetscher HR, Brockhaus M, Dembic Z, Gentz R, Gubler U, Hohmann HP, Lahm H W van Loon APGM, Pan YCE, Schlaeger EJ, Steinmetz M, Tabuchi H, Lesslauer W: Two distinct tumor necrosis factor receptors - members of a new cylokine receptor gone family; in MacIean (ed): Oxford Survey on Eucaryotic Genes. Oxford, Oxford University Press. 1991. 24 Hohmann H, Remy R, 8rockhaus M, van Loon APGM: Two different cell types have different major receptors for human tumor necrosis factor (TNF alpha). I Biol Chem 1989;264:14927. 25 Brockhaus M, Schoenfeld H J, Schlaeger EJ, Hunziker W, Lesslauer W, Loetscher HR: Identification of two types of tumor necrosis factor receptors on human cell lines by monoclonal antibodies. Proc Natl Acad Sci USA 1990;87:3127. 26 Loetscher HR, Schlaeger EJ, Lahm HW, Pan YCE, Lesslauer W, Brockhaus M: Purification and partial amino acid sequence analysis of two distinct tumor necrosis factor receptors from HL60 cells. J Biol Chem 1990;265:20131, 27 Loetseher H: Purification and structural properties of two distinct TNF receptors on h u m a n cells; in Aggarwal, Vilcek (eds): Tumor necrosis factors - structure, function and mechanism of action; Marcel Dekker, New York, ! 991.
28 Mallett S, Barclay AN: A new superfamily of cell surface proteins related to the nerve growth factor receptor, lmmunol Today 1991;12: 220. 29 Rogers S, Wells R, Rechsteiner M: Amino acid sequences common to rapidly degraded proteins. The PEST hypothesis. Science 1986;234: 364. 30 Seckinger P, Isaaz S, Dayer JM: A human inhibitor of tumor necrosis factor alpha. J Exp Med 1988;167: 1511. 31 Seckinger P, Isaaz S, Dayer JM: Purification and biologic characterisation of a specific tumor necrosis factor alpha inhibitor. J Biol Chem 1989;264: t 1966. 32 Olsson I, Lantz M, Nilsson E, Peetre C, Thysell H, Grubb A, Adolf G: Isolation and characterisation of a tumor necrosis factor binding protein from urine. Eur J Haematol 1989;42:270. 33 Peetre C, Thyseel H, Grubb A, Olsson I: A tumor necrosis factor binding protein is present in human biological Huids. Eur J Haematol 1988; 41;414. 34 Engelmann H, Aderka D, Rubinstein M, Rotman D, Wallach D: A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem 1989;264:11974. 35 Engelmann H, Novick D, Wallach D: 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 1990;265:1531. 36 Olsson I, Lantz M, Nilsson E, Peetre C, ThyseU H, Grubb A, Adolf G: Isolation and characterisation of a tumor necrosis factor binding protein from urine. Eur J Haematol 1989;42:270. 37 Seckinger P, Vey E, Turcatti G, Wingfield P, Dayer JM: Tumor necrosis factor inhibitor. Purification, NH2-terminal amino acid sequence and evidence for anti-inflammatory and immunomodulatory activities. EurJ Immuno11990;20:1167.
38 Digel W, Por'zsolt F, Lesslauer W, Brockhaus M: Evidence for soluble receptors for tumor necrosis factor in the serum of cancer patients high levels in B-cell malignancies, submitted. 39 Gatanaga T, Hwang C, Kohr W, Capuccini F, Lucci IA, Jeffes EWB, Lentz R, Tomich J, Yamamoto RS, Granger GA: Purification and characterisation of an inhibitor (soluble tumor necrosis factor receptor) for tumor necrosis factor and lymphotoxin obtained from the serum ultrafiltrates of human cancer patients. Proc Natl Acad Scl USA 1990;87: 8781. 40 Lesslauer W, Tabuchi H, Gentz R, Schlaeger EJ, Brockhaus M, Grau G, Piguet PF, Pointaire P, Vassalli P, Loetscher H: Bioactivity of recombinant human TNF receptor fragments. J Cell Biochem 1991; 15F(suppl): I 15. 41 Loetscher HR, Gentz R, Zulauf M, Lustig A, Tabuchi H. Schlaeger E J. Brnekhaus M, Gallati H, Manneberg M, Lesslauer W: Recombinant 55kDa TNF receptor. Stoichiometry of binding to TNFalpha and TNF'beta and inhibition o f T N F activity~ J Biol Chem, in press. 42 Hohmann HP, Remy R, Poeschl B, van Loon APGM: Tumor necrosis factors-alpha and -beta bind to the same two types of tumor necrosis factor receptors and maximally activate the transcription factor NK-kB at low receptor occupancy and within minutes after receptor binding. J Biol Chem 1990;265:15183. 43 Engelmann H, Holtmann H, Brakebusch C, Avni SY, Sarov I, Nophar Y, Hadas E, Leitner O, Wallach D: Antibodies to a soluble form of a tumor necrosis factor (TNF) receptor have TNF-like activity. J BiN Chem 1990;265:14497. 44 Hibi M, Murakami M, Saito M, Hirano T, Taga T, Kishimoto T: Molecular cloning and expression of an IL-6 signal transducer, gpl30. Cell 1990;63:1149. 45 Kobayashi N, Hamamoto Y, Yamamoto N, Ishii A, Yonehara M, Yonehara S: Anli-FAS monoclonal antibody is cytocidal to human immunodeficiency virus-infected ceiis without augmenting viral replication. Proc Natl Acad Sci USA 1990; 87:9620.
89
46 Brouckaert PG, Everaerdt B, Libert C, Takahashi N, Fiers W: Species specificity and involvement of other cytokines in endotoxic shock action of recombinant tumor necrosis factor in mice. Agents Actions 1989;26: 196. 47 Hohmann HP, Brockhaus M, Baeuerle PA, Remy R, Kolbeck R, van Loon APGM: Expression of the types A and B tumor necrosis factor (TNF) receptors is independently regulated and both receptors mediate activation of the transcription factor NF-kB. TNFalpha is not needed for induction of a biological effect via TNF receptors. J Biol Chem 1990;265:22409. 48 Gehr G, Gentz R, Loetscher HR, Brockhaus M, Tabuchi H, Lesslauer W: TNF receptors alpha and beta (TNFRalpha, TNFRbeta) in human mononuclear cell activation, submitted. 49 Thoma B. Grell M, Pfizenmaier K, Scheurich P: Identification of a 60 kD tumor necrosis factor (TNF) receptor as the major signal transducing component in TNF responses. J Exp Med 1990;172:1019. 50 Meichle A, Schuetze S, Hensel G, Brunsing D, Kroenke M: Protein kinase C-independent activation of nuclear factor kB by tumor necrosis factor. J Biol Chem 1990;265:8339.
51 Brenner DA, O'Hara M, Angel P, Chojkier M, Karin M: Prolonged activation ofjun and collagenase genes by tumor necrosis factor alpha. Nature 1989:337:661. 52 Harada H, Fujita T, Miyamoto M, Kimura Y, Maruyama M, Furia A, Miyata T, Taniguchi T: Structurally similar but functionally distinct factors IRF-1 and IRF-2 bind to the same regulatory elements of IFN and IFN-inducible genes. Cell 1989; 58:729. 53 Zhang Y, Lin JX, Vilcek J: Interleukin 6 induction by tumor necrosis factor and interleukin 1 in human fibroblasts involves activation of a nuclear factor binding to a kB like sequence. Mol Cell Biol 1990;10: 3818. 54 Ray A, Sassone-Corsi P, Sehgal P: A multiple cytokine- and second messenger-responsive element in the enhancer of the human interleukin-6 gene. Similarities with c-fos gene regulation. Mol Cell Biol 1989;9: 5537. 55 Akira S, Isshiki H, Sugita T, Tanabe O, Kinoshita S, Nishio Y, Nakajima T, Hirano T, Kishimoto T: A nuclear factor for IL-6 expression (NFIL6) is a member ofa C/EBP family. EMBO J 1990;9:1897. 56 Schuetze S, Scheurich P, Pfizenmaier K, Kroenke M: Tumor necrosis factor signal transduction. Tumor specific serine phosphorylation of a 26 kDa cytosolic protein. J Biol Chem 1989;264:3562.
57 Donato NJ, Gallick GE, Steck PA, Rosenblum MG: Tumor necrosis factor modulates epidermal growth factor receptor phosphorylation and kinase activity in human tumor cells. Correlation with cytotoxicity. J Biol Chem 1989;264:20474. 58 Marino M, Pfeffer LM, Guidon PT, Donner DB: Tumor necrosis factor induces phosphorylation of a 28 kDa mRNA cap-binding protein in human cervical carcinoma cells. Proc Natl Acad Sci USA 1989;86: 8417. 59 Kohno M, Nishizawa N, Tsujimoto M, Nomoto H: Mitogenic signalling pathway of tumor necrosis factor involves the rapid tyrosine phosphorylation of 41'000 M~ and 43"000 Mr cytosol proteins. Biochem J 1990; 267:91. 60 Shiroo M, Matsushima K: Enhanced phosphorylation of 65 and 74 kDa proteins by tumor necrosis factor and interleukin 1 in human peripheral blood mononuclear cells. Cytokine 1990;2:13. 61 Bird TA, Saklatvala J: IL-I and TNF transmodulate epidermal growth factor receptors by a protein kinase C-independent mechanism. J lmmunol 1989;142:126. 62 Hohmann HP, Remy R, Schneidereit C, van Loon APGM: Cyclic AMP independent activation of the transcription factor NF-kB in HL60 cells by tumor necrosis factors alpha and beta. Mol Cell BioI [991;11: 2315.
90
Rothe/Gehr/Loetscher/Lesslauer
TNF Receptors
