CA2490080A1 - Modulator of tnf/ngf superfamily receptors and soluble oligomeric tnf/ngf superfamily receptors - Google Patents

Abstract

The present invention generally concerns novel proteins which bind to the intracellular domains of the p55 and p75 TNF-Rs and the Fas-R, which are capable of modulating the function of the p55 and p75 TNF-Rs and the Fas-R, and the DNA
sequences which encode them. The present invention also concerns new soluble oligomeric TNF-Rs, oligomeric Fas-Rs and oligomeric receptors having a mixture of TNF-Rs and Fas-Rs. In addition, the present invention concerns methods of preparation and uses of all of the aforementioned.

Description

MODULATOR OF TNF/NGF SUPERFAMILY RECEPTORS AND SOLUBLE
OLIGOMERIC TNF/NGF SUPERFAMILY RECEPTORS
Field of the Invention The present invention is generally in the field of receptors belonging to the TNF/NGF
superfamily of receptors and the control of their biological functions. The T'I~'FINGF superfamily of receptors includes receptors such as the p55 .and p75 tumor necrosis factor receptors ('TI~tF-Rsl and the FAS ligand receptor (also called FAS/APOI or FAS-R and hereinafter will be called FAS
R) and others. More specifically, the present invention concerns novel proteins which bind to the intracellular domains (IC) of the p55 and p7~ TNF-Rs and the Fas-R, (these intracetlular domains designated p55IC, p75IC and Fas~IC, respectively) and which novel proteins are capable of modulating the function of the p:~ and p?5 TNF-Rs and the Fas-R. One of the proteins capable of binding the p55IC of the imact p~5-TIr'F R is the p55IC itself in the form of a p55IG molecule or a portion thereo>~ such as for example. the so-called 'death domain' (DD) of the p55IC. Thus, the present inversion also concerns new T's1F-associated effects that can be induced in cells in a ligand (TNF)-independent fashion by the intracellular domain of the p55 TNF-R
(p55IC) or portions thereof. The present invention also concerns the preparation and uses of these novel p55 and p75 TNF-R-binding proteins, and Fas-R binding proteins, referred to herein as pSSIC-, p75IC- and Fas-IC- binding proteins.
In another aspect, the present imrention also concerns new soluble oligomeric TNF-Rs, oligomeric FAS-Rs and oligomeric receptors having a mixture ofThl'F-Rs and I~AS-Rs, their uses, and methods for the production thereof.
BackøroLad of the Invention and Prior Art , Tumor Necrosis Factor (TNF-a) and Lyrnphotoxin (TNF-Vii) (hereinafter, TNF, refers to both TNF-a and TNF-p) are multifunctional pro-inflammatory cytolines formed mainly by mononuclear phagocytes, which have many effr'ects on ceps (Wallach, D. (1986) in : Interferon 7 (Ion Dresser, ed.), pp. 83-I22, Academic Press, London; and Beutler and Cerami (1987)). Both TNF-cc and TNF-~i initiate their effects by binding to specific cell surface receptors. Some.of the effects are likely to be beneficial to the organism : they may destroy, for example tumor cells or virus infected cells and augment arsibactaial activities of granulocytes. In this way, ;TNF
contributes to the defense of the organism against tumors and infectious agents and contributes to the recovery from injury. Thus. TNF .can be used as an anti-tumor agent in which application it binds to its receptors on the surface of tumor cells and thereby initiates the events leading to the death of the tumor cells. TNF can also be used as an anti-infectious agent.

WO 95!31544 PCT/US95/05854 However, bole. 5?v'F-a and TNF-G also have deleterious er'focts. There is ev idence that over-production of T~; -a can play a major pathogenic role in several diseases. Thus, effects of TNF-a, primarily on t: vasculature, are now known to be a major cause for symptoms of septic shock (Tracey et al., 1~SG). In some diseases, Tiv'Ir' may cause excessive loss of weight (caehexia) by suppressinb artivit:a of adipocytes and by causing anorexia, and TNF-a was thus called cachetin. h was also ;:eseribed as a mediator of the damage to tissues in rheumatic diseases (Beutler and Cerami, 1S?) and as a major mediator of the damage observed in graft-versus-host reactions (Piquet et a:., 1987). In addition, TVF is known to be involved in the process of inflammation and in mt~tv other diseases.
Two distinct, independently ecpressed, receptors, the p~~ and p7S T~TF-Rs, which bind both Th(F-a and T~If-G specifics!!}~, initiate and/or mediate the above noted biological effects of TNF. These two receetors have structurally dissimilar intracellular domains suggesting that they signal differently (See Hohmann et al., 1989; Engelmann et al., 1990;
Brockhaus et al., 1990;
Leotscher et al., I99C'. Schall et al., 19.90; Nophar et aL, I990; Smith et al., 1990; and Heller et al.. 1990). I-3owever, the cellular mechanisms, for example, the various proteins and possibly other factors, which are invoiced in the inuacellular signaling of the p55 an p i 5 TIv'F-Rs have yet to be elucidated {as set forth herein below. there is described for the first time, new proteins capable of binding to the p?SIC and p55 IC). It is this intracellular signaling, which occurs usually after the binding of the Iigand, i.e. TNF (a or ~), to the receptor, that is responsible for the commencement of the cascade of reactions that ultimately result in the observed response of the cell to TNF.
As regards the above mentioned cytocidal effect of TNF, in most cells studied so far, this effect is triggered mainly by the p55 TNF-R Antibodies against the e~.-tracellular domain {ligand binding domain) of the p55 TNF-R can themselves trigger the cytocidal effect (see EP 413486) which correlates with the effectivitt~ of receptor cross-linking by the antibodies. believed to be the first step in the generation of the intracellular signaling process. Further, mtvational s~.:~dies (Brakebusch et al., 199; TartagIia et al., 1993) have shown that the biolo~rical function of the pS5 TNF-R depends on the intcgrit;~ of its intracellular domain, and accordingly it has been suggested that the initiation of intracellular signaling leading to the cytocidal effect of TNF occurs as a consequence of the association of two or more intracellular domains of the p55 TNF-R.
Moreover, TNF (a and Vii) occurs as a homotrimer and as such has been sugbested to induce intracellular signaling va the p55 7'Nf-R by way of its ability to bind to and to cross-link the receptor molecules, i.e. cause receptor aggregation. Herein below there is described how the pSSIC and pSSDD can self associate and induce, in a figand-independent fashion, T';~F-associated effects in cells.
Another member of the TNFINGF supecfamily of receptors is the FAS receptor (FAS-R) which has also been called the Fns antigen, a cell-surface protein expressed in various tissues .aid sharing homology with a numbs of cell-surface receptors including TNF-R and NGF-R. ?he FAS-R mediates cell death in the form of apoptosis (Itoh et al., 1991 ), and appears to serve as a negative selector of autoreactive T cells, i.e. during maturation of T cells, FAS-R mediates the apoptopic death of T cells recognizing self antigens. It has also been found that mutations in the WO 95!31544 FAS-R gene (Ipr) cause a lymphoprofiieration disorder in mice that resembles the human autoimmune disease systemic Iupus erythematosus (SLE) (Wacanabe-Fukunaga et al.. 1992). The ligand for the FAS-R appears to be a cell-surface associated molecule carried by, amongst others, killer T cells (or cytotoxic T lymphocytes - CTLs), and hence when such CTLs contact cells carrying FAS-R, they are capable of inducing apoptopic cell death of the FAS-R-carrying cells.
Further, a monoclonal antibody has been prepared that is specific for FAS-R, this monoclonal antibody being capable of inducing apoptopic cell death in cells carrying FAS-R, including mouse cctls transformed by cDNA encoding human FAS-R (Itoh et al., 1991).
It has also been found that various other normal cells, besides T lymphocytes, e~cpress the FAS-R on their surface and can be killed by the triggering of this receptor.
Uncontrolled induction of such a killing process is suspected to contribute to tissue damage in certain diseases, for example, the destruction of liver cells in acute hepatitis. Accordingly, finding ways to restrain the cytotoxic activity of F AS-R may have therapeutic potential.
Conversely, since it has also been found that certain malignant cells and HIV-infected cells carry the FAS-R on their surface, antibodies against FAS-R, or the FAS-R
licrand, may be used to trigger the );AS-R mediated cytotoxic effects in these and Lhereby provide a means for combating such malignant cells or HIV-infected cells (see Itoh et al., 1991 ). Finding s-et other ways for enhancins~ the cytotoxic activity of FAS-R may therefore also have therapeutic potential.
It has been a long felt need to provide a u~ay for modulating the cellular response to TNF
a or ~) and FAS-R ligand, for a:ampte, in pathological situations as mentioned above, where TNF or FAS-R ligand is over-expressed it is desirable to inhibit the TNF- ar FAS-R lieand-induced cytocidal effects, while in other situations, e.g. wound healing applications, it is desirable to enhance the TNF effect, or in the case of FAS-R in tumor cells or HI'f-infected cells it is desirable to enhance the FAS-R mediated eB'ect.
A number of approaches have been made by the present inventors (see for example, European Application Nos. EP 186833, EP 308378, EP 398327 and EP 412486) to regulate the deleterious effects of TNF by inhibiting the binding of TNF to its receptors using anti-TNF
antibodies or by using soluble TNF receptors (being essentiaDy the soluble ea~tracellular domains of the receptors) to compete with the binding of TNF to the cell surface-bound TNF-Rs. Further, on the basis that TNF-binding to its receptors is required for the TNF-induced cellular effects.
approaches by the present inventors (see for example EPO 568925) have been made to modulate the TIvF effect by modulating the actt<~ity of the Tlv'F-Rs. Briefly, EPO
X68925 relates to a method of modulating signal transduction and/or cleavage in TNF-Rs whereby peptides or other molecules may interact either with the receptor itself or with effector proteins interacting with the receptor, thus modulating the normal functioning of the TNF-Rs. In EPO 568925 there is desetibcd the construction and characterization of i~arious mutant p55 TNF-Rs, having mutation in the extraceUular, transmembranal, and intracellular domains of the p55 TNF-R. In this way regions within the above domains of the p55 TNF-R were identified as being essential topthe functioning of the receptor, i.e. the binding of the Iigand (?NF) and the subsequent signal ttansduction and intracellular signaling which ultimately results in the observed TNF-effect on the cells. Further, there is also described a number of approaches to isolate an~
identify proteins, peptides or other factors which are capable of binding to the various regions ir. the above domains of the TNF-R, which proteins, peptides and other facxors may be involu:d in regulating or modulating the activity of the TNF-R. A number of approaches for isolating ar d cloning the DTIA
sequences encoding such proteins and peptides; for constructing expression vectors for the production of these proteins and peptides; and for the preparation of antibodies or fr$emcnts thereof which interact with the TNF-R or with the above proteins and peptides that bind various reaons of the TNF-R, are also stt forth in EPO 568925. However, no description is made in EPO
X68925 of the actual proteins and peptides which bind to the intracvlluiar domains of the TNF-Rs . 1 G (e.g. p55 TNF-R), nor is any description made of the yeast two-hybrid approach to isolate_ar~d identify such prottins or peptides which bind to the irnracellular domains of TNF~Rs. Similarly, heretofore there has been no disclosure of proteins or peptides capable of binding the intracellular domain of FAS R. ._ Thus, when it is desired to inhibit the effect of T'._v'F, or the FAS-R
:igand, it would be l ~ desirable to decrease the amaum or the activity of ?NF Rs or FAS-R at the cell surface, while.arl increase in the amount or the activity of TNF-Rs or FA,S-R would be desired when an enhanced TNF or FAS-R tigand effect is sought. Ta this end the promoters of both the p5~ T~1'F-R and.~the p75 T13F-R have recently been sequenced and analyzed by the present inventors and a number of kcs~ sequence motifs have been found that are specific to various transcription regulating factors, 2C~ and as such the expression of these TNF Rs can be controlled at their promoter level, ..i:e.
itdu'bition of transcription from the promoters for a decrease in the number of receptors, and an enhancement of transcription from the promoters for an increase in the number of receptors.
Ca:responding studies eaneernirrg the control of FAS-R at the level -of-.the 25 promoter of the FAS-R gone have yet to be reported.
Further, it should also be mentioned that, while it is known that the tumor necrosis factor ' (T'N~ receptors, and the structurally-rtlated receptor FAS-R; trigger in cells, upon stimulation by leukocyte-produced ligands, destructive activities that Lead to their own demise, the mechanisms ef this triggering are still little understood. Mutational studies indicate that in FAS~R and the:p35 30 TNF receptor (p55-R) signaling for cytotoxiasy imrolve distinct regions within their intracellular domains (Brakebusch et al,, 1992; Tartaglia et al., 1993; hoh and Nagata, 1993). These.repons (the 'death dotnains~ have sequence similarity. The 'death domains' of both FAS-R and.:p5,5~R
tend to self assoaate. Their self association apparently promotes that receptor aggregation fyvhich is necessary for itutiation of signaling (as set forth herein below, as well as Song at al.,, ~199.i;
3 5 Wallach et aL, 1994; Boldin et al., ~ I 995) and at high Levels of receptor expression can result .in trigsering of ligand~ndependern signaling (as set forth herein below, and Boldin et al., 199,). .
Thus, prior to the present invention, there have not been provided proteins which may regulate the effect of Iigands belonging to the TN'1~INGF superfamily, such as the TNF or.FAS-R
Iigand effect on ~eclls, by mediation of the intracellular signaling process, which signaling ;is 40 probably governed to a Large extent by the intracellular domains (ICs) of the receptors below to the ThIFINGF superfamify of receptors, such as those of the TNF-Rs, i c.
the p55 and p?~
T:~IF-R intracellular domains (p55IC and p75IC, respectively), as well as the F AS-IC.
Accordingly, it is one aim of the invention to provide proteins u~h:ch are capable of bindinb to the intracellular domains of the TNF-Rs and FAS-R which proteins are presentlw believed to be involved in the intracellular signaling process initiated by the binding of TNF to its receptors, or the bindinb of FAS ligand to its receptor.
Another aim of the invention is to provide antagonists (e.g. antibodies) to these intracellular domain-binding proteins (IC-binding proteins) which may be used to inhibit the si~,:naling process, when desired, when such IC-binding proteins are positive signal effectors (i.e.
lu induce siunaling), or to enhance the signaling process, when desired, when such IC-bindin_~
proteins are neLative signal effectors (i.e. inhibit signaling), Yet another aim of the incemion is to use such IC-bindinb proteins to isolate and characterize additional proteins or factors, which may, for e:cample, be involved further downstream in the signaling process, and/or to isolate and identify othe:
receptors further 1~ upstream in the signalins process to which these IC-binding proteins bind (e.g.~other T'I~rF-Rs or related receptors ), and hence, in whose function the IC-binding proteins are also involved.
Moreover, it is an aim of the present invention to use the above-mentioned IC-binding proteins as antigens for the preparation of polyclonal andlor monoclonal antibodies thereto. The antibodies, in turn, may be used for the purification of the new IC-binding proteins from difl'crent 2G sources, such as cell extracts or transformed cell lines.
Furthermore, these antibodies may be used for diagnostic purposes, e.g. for identifying disorders related to abnormal functioning of cellular effecu mediated by receptors belonging to the TNFINGF receptor superfamily. .
A further aim of the invention is to provide pharmaceutical compositions comprising the 2~ above IC-binding proteins, and pharmaceutical compositions comprising the IG-binding protein amaganists, for the treatment or prophylaxis of TNF-induced or FAS ligand-induced conditions, for example, such compositions can be used to enhance the TI\'F or F AS Iigand effect or to inhibit the 'fNF or FAS ligand effect depending on the above noted natwe of the IC-binding protein or antagonist thereof contained in the composition.
30 Moreover, in accordance with another aim of the present invention, there is disclosed other ways for eliminating or antagonizing endogenously formed or exogenously administered TNF or FAS-R ligand, by the use of soluble oligomeric TNF-Rs, oligomeric FAS
Rs, or oligomers being a mixture of TNF-Rs and FAS-Rs. In this respect it should be mentioned that one attempt in this direction was the isolation and recombinant production of a TIFF Binding Protein 35 called TBP-1 which was shown to be able to antagonize the effects of TNF.
This a~rtagonistn was determined both by measuring reduction of the cytotoxic activity of TNF', as well as by measuring interference of ~NF binding to its receptors (EP 308 378). ?BP-I was shown to protect cells from TNF toxiaty at concentrations of a few nano~ams per ml and to interfere v~~ith the binding of both 'rNF-a and TNF-p to cells, when applied simultaneously with these cytoltines. Further 40 examination of the mechanism by which THP-I functions revealed that TBP-I
does not interact G
with the target cell, but rather blocks the function of TNF by binding 'I1-F
specifically, thus competing for TNF with the TNF receptor.
Consequently, with a different purification technique, the presence of two active components was found : one, TBP-I, and also a second TIv'F-binding protein which we called TBP-II (first described in EP 398327). Both proteins provide protection zgainst the ire vitro cyZOCidal effect of TNF and both bind T~1F-p less effectively than TNF-c:
Although in SDS
PAGE analysis the two proteins, TBP-I and TBP-II, appeared to have a wen-similar molecular size, they could clearly be distinguished from cach other by Lack of immunolo=
cal cross reactivity, differing N-terminal amino acid sequences and differing amino acid composition.
Fiowcver, the above noted earlier soluble TNF binding proteins are manomeric and being capable of binding only one monomer of the TNF homotrimer, the natural Iigand, which still permits TVF activity (i.e. incomplete neutralization) by virtue of the fNF
stil: having two active monomers unbound by the TIvTr binding proteins. Further, heretofore there ha:
been no disclosure of soluble FAS-Rs (soluble FAS-R ligand binding proteins) capable of binding to 1~AS-R ligand 1~ which is known to be a homotrimeric, cell-surface associated molecule.
A so-called 'death domain' of the pss-IC (Tartaglia et al., 1993) hzs been disclosed, but did not show, in accordance with the present invention, that the p55-IC and the 'death domain' thereof self associates, this self association being primarily responsible for the si'naling leading to induction of cell c5~totoxis. Moreover, this publication is silent on the possibility of produc'mg the soluble, oligomeric TNF-Rs, or the soluble, oligomeric Fas-Rs, or mixed oligomeric thereof nor does it disclose other TNF-associated effects induced by the p55-IC or portions thereof, e.g. IL-8 gene expression induction, all of the present invention. Likewise, another publication, published after the date of the present invention, disclosed the aggegation (i.e, self association) ability of the p55-IC, but did not relate, as noted above, to the usage thereof to prepare soluble, oligomeric TNF-Rs or Fas-Rs nor to the other TNF-associated effects induced in a lis;a: d-indepo~a~t manner by the p55-IC or portions thereof according to the invention Summary of the Invention In accordance with the present invention, we have found novel proteins which are capable of binding to either the intracellular domain of the p55 T'!VF-R (the pSSIC-binding proteins), of the p75 TNF-R (the p75IC-binding proteins), and of the FAS-R (the FAS-IC-binding proteins).
Thcsc p55IC-, p75IC- and FAS-IC- binding proteins may act as mediators or modulators of the T.NF or FAS-R liband effect on cells by way of mediating or modulating the intracellular signaling process which usually occurs following the binding of TNF to the p55 andlor p75 TNF-R, or the binding of the FAS-R Iigand at the cell surface. Further, it has been surprisingly and unexpectedly found that the p55IC and FAS-IC are capable of self association and that fragments of the p55IC-and FAS-IC are similarly capable of binding to the p55 IC, particularly the so-called 'death domains (DD) within the ICs of these receptors, i.e. the p55DD and FAS-DD.
Thus, p55 IC and FAS-IC and their fragnents also represent proteins capable of binding to the p55IC and FAS-IC
and hence may be modulators of the 'f~TF or FAS-R ligand effect on cell s.

WO 95!31544 PCT/US95/05854 Furthermore, the nature of the binding of one of the novel proteins of the invention, the herein designated ~5. i 1 protein, to the intracellular domain of p55-TNF-R
has been more fully elucidated (see Example 1 ), Moreover, in another aspect, the present invention is based on the finding that the S intracellular domain of the p55 TNF receptor (p55-IC), a region contained therein, the so-called p55-IC 'death domain', the intracellular domain of the FasIAPOI receptor fFas-1C), and a re~~ion contained therein, the so-called Fas-IC 'death domain' are capable of self association.
Accordingly, it is possible to construct by standard recombinant DNA
techniques, a soluble, oligomeric TNF receptor being a fusion product, containing at least two extracellular domains of 14 a TNF receptor at its one end, and at its other end at least two of the above noted self associating invacellular domains or portions thereof, which self associate to provide an olioomer having at Ieast two such fusion products linked together. Such a soluble, oligomeric T1'F-R is thus capable of binding two monomers of the naturally-occurrinb TTv'F homotrimer, and as such effectively neutralizes T'fF activity. The neutralization of TNF activity being desirable in all of the above 15 mentioned conditions wherein TNF is overproduced endogenously or is administered exoeenously in high doses resultin' in undesirable side effects. Further, the effective binding of TNF by the soluble, oligomeric receptors of the invention may also serve to allow for the binding of exoeenously added TI~rF and its subsequent desired slow-release in conditions where TNF is administered for its beneficial effects, e.g. in tumor therapy. Likewise, it is also poss'bIe to 20 construct by standard rerombittant DNA techniques an oLigomeric FAS-R being a fusion product, containing at least two exvaceUular domains of a FAS-R at its one end, and at its other end at least two of the above noted self associating intracellular domains or portions thereof, which self associate to provide an oligomer having at least two such fusion products linked together. Such an oligomcric FAS-R is thus capable of binding two monomers of the naturally occurring FAS R
25 ligand homotrimer, and as such effectively neutralizes FAS-R ligand actiZ~iy The n°~itralizatio~. of FAS-R ligand activity being desirable in all of the above mentioned conditions where excess amounts thereof are associated with undesirable side effects. In a similar fashion.. and in view of tecent reports indicating a possible associating between 'TNF and FAS-R ligand-induced effects on cells and hence also a possible association, geo~aphically at the cell surface where they attach 30 to their receptors, it is also possible to construct by standard recombinant DNA techniques a mixed oligomeric receptor having specificity for both Ti'JF and FAS-R
lic_:and. Such a mixed oligomer would be a mixture ~of the above noted fusion products containing at least one extracellular domain of a TNF-R and at least one extracellutar domain of a FAS-R at its one end, and at its other end at least two of the above mentioned self associating intracellular domains or 3 S portions thereof, which self associate to provide a tnixcd oligomer having at least two such fusion t -products linked together. Such a mined oligomer is thus capable of binding at least one monomer of TI~tF and one monomer of F AS-R ligand at the same time, thereby reducing or effectively neutralizing the TNF and FAS-R ligand activities at the cell surface in conditions, as noted above where excess amounts of these two cytokines are associated with undesirable cellular effects. As 40 noted above, the FAS-R ligand is usually cell-surface-associated, and recent reports also describe WO 95131544 PCTlLjS95105854 s cell-surface-associated forms of TNF. Hence, these mixed TNF-R~FAS-R oli~omers are especially useful for neutralization of TNF and FAS-R ligand activities at the cell surface.
Accordingly, the present invention pro~~ides a D~1.4 sequence enco3in_ a protein capable of binding to one or more of the intracellular domains of one or more re;.cptors belonging to the tumor necrosis factorinen~e growth factor (TNFlNGF} superfamily of receptors.
_ In particular. the present invention provides a DNA sequence selected from the group consisting of (a) a cDNA sequence derived from the coding region of a native TNF-R
intracellular domain-bindinb protein;
(b) DNA sequences capable of hybridization to a DNA of (a) under moderately stringent conditions and which encode a biologically active TI~TF'-R intracellular domain-binding protein; and (c) DNA sequences which are degenerate as a result of the genetic code to the DNA
sequences defined in (a) and (b) and which encode a biologically active TNF-R
I 5 intracellular domain-binding protein.
The present invention also provides a DN'A sequence selected from the group consisting of (a) a cDNA sequence derived from the coding region of a native F AS-R
intracellular domain-binding protein;
(b) DNA, sequences capable of hybridization to a cDNA of (a) under moderately stringent conditions and which encode a biolopeally active FAS-R intracellular domain-binding protein; and (c) DNA sequences which are degenerate as a result of the genetic code to the DNA
sequences def ned in (a) and (b) and which encode a biologically active FAS-R
intracellular domain-binding protein.
In embodiments of the present invention the DNA sequences encode p55 TNF-R, p75 TNF-R and FAS R irnraceUular domain-binding proteins, such as those encoding the herein designated proteins 55.1, 55.3, 55.I I, 75.3, ?5.16, F2, F9 and DD11.
The present invention also provides a protein or analogs or derivatives thereof encoded by any of the above sequences of the invemion, said proteins, analogs and derivatives being capable of binding to one or more of the intracdlular domains of one or more TNF-Rs or FA'S ;R.
Embodiments of this aspect of the invention include the herein designated proteins 55,1, 55.3, 55.11, 75.3, 75.16, F2, F9 and DDI 1, their analogs and their dctivatives. .
Also provided by the present invention are vectors encoding the above proteins of the invention, which contain the above DNA sequences of the invention,.these vectors being capably' of being expressed in suitable cukaryotic or prokaryotic host cells;
transformed eukaryotic' or prokaryotic host cells containing such vectors; and a method for produrang the proteins, analogs or derivatives of the im~esxtion by growing such transformed host cells under conditions suitable for the expression of said protein, analogs or derivatives, efr'ecting post-translational modifications of said flrotein . .as necessary for obtention of said protein and extracting said expressed protein analogs or derivatives from the culture medium of said transformed cells c:
from cell extracts of said transformed cells.
In another aspect. the present invention also provides antibodies o: active derivatives or fragments thereof specific to the proteins, analogs and derivatives thereof, of the invention.
By yet another aspect of the invention, there are provided various uses of the above DNA
sequences or the proteins which they encode. according to the invention, which uses include amongst others (i) a method for the modulation of the TIv'F or FAS-R lisand ~ficct on calls carrying a TNF-R or a FAS-R, comprising treating said cells with one or mc~rc proteins, analogs or derivatives selected from the group consisting of the proteins, analogs and derivatives, according to the invention, and a protein being the p55IC, p~SDD. FAS-IC or FAS-DD, analogs or derivative: thereof, all of said proteins being capable of binding to the intracellular domain and modulatinL the activity of said T: TF-R or FAS-R
wherein said treating of the cells comprises introducing into said cells said one or more proteins.
analogs or derivatives in a farm suitable for intracellular administration or introducing into said cells, in the form of a suitable expression vector, the DNA sequence encoding said one or more proteins, analogs or dcrivativ es ;
(ii) a method for modulating the TI''F or FAS-R li?and effect on cells carrying a TNF
R or a FAS-R comprising treating said cells with antibodies or active derivatives or 20_ fragments thereof according to the invar~tion;
(iii) a method for modulating the TNF or FAS-R ligand effect on cells carrying a TTIF-R or FAS-R comprising treating said cells with an oligonueleotide sequence encoding an antisense sequence of at least part of the sequence according to the invention, or encoding an antisense sequence of the p55IC, pS~DD, FAS-1C, or FAS-DD
sequence.
said oligonucleotide sequence being capable of blocking the e;cpression of at least one of the TNF R or FAS-R intraccUular domain binding proteins;
(iv) a method for modulating the TNF or FAS-R iigand efrect on cells carrying a TNF-R or FAS-R comprising (a) constructing a recombinant animal virus vector carrying a sequence encoding a viral surface protein that is capable of binding to a specific cell surface receptor and a sequence selected from an oligonucleotide sequence encoding an antisrnsc sequence of at least part of the sequence according to the invention and an oligonucleotide sequence encoding an antisensc sequence of the p~3IC, p55DD, FAS-IC, or FAS-DD sequence, said oligonucleotide sequence being capable of blocL-ing the expression of at least one of the TNF-R or FAS R intracellular domain binding proteins v~~hen introduced into said cells by said virus,; and (b) infecting said cells with said vector of (a).
(v) a method for modulating the TIv'F or FAS-R tigand rffect on cells carr~ring a TNF
R or a FAS-R, comprising treating said cells with a suitable vector encoding a ribozymc X10 having a sequence specific to a sequence selected from an mRI~TA sequence encoding a WO 95131544 PCTlUS95105854 IO
protein, analog: or derivative of the invention and an mR,'A sequer:e encodin~T the p55IC.
p55DD, F.45-IC or FAS-DD, said :ibozyme sequence capable of interacting with said mRNA sequence and capable of cleavin' said mRNA sequence resulting in the inhibition of the expression of the protein, anaioe or dert'~ativc of the invention or of the e~cpression of the p:SIC, p55DD, FAS-IC or FAS-DD.
(vi) a method for treating tumor cells or HIV-infected cells, or other diseased cells, compri sing (a) constructing a recombinant animal virus vector carrying a sequence encoding a viral surface protein that is capable of binding to a tumor cell surface receptor or HIV
infected cell surface receptor or is capable of binding to another cell surface receptor of other diseased cells and a sequence selected from a seauence according to the invention encoding a protein, analog or derivative of the invention and a sequence encoding the p55IC, p~SDD, FAS-IC, FAS-DD, or a biologically active analot or derivative thereof.
said protein, analog or derivative ef the invention, p5 SIC, p5 f DD, FAS-JC, FAS-DD.
IS analob or, derivative, when expressed in said tumor cell or HIS'-infected cell, or other diseased cell being capable of killing said cell; and (b) infecting said tumor cells or HIL'-infected cells or other infected cells with said vector of (a).
(vii) a method for isolating and identifying proteins, factors or receptors capable o binding to the intracellular domain binding proteins according to the invention, comprisinc applying the procedure of a~'miry chromatography in which said protein according to the invention is attached to the af~nnity chromatography matrix, said attached protein is brought into contact with a cell extract and proteins, factors or receptors from cell extract which bound to said attached protein arc then eluted, isolated analyzed;
(W i) a method for isolating and identifying proteins, capable of bindini to the intracellular domain binding proteins according to the invention, comprising applying the yeast two-hybrid procedure in which a sequence encoding said intracellular domain binding protein is carried by one hybrid vector and a sequence from a cDNA or genomic DNA library is carried by the second hybrid vector, the vectors then being used to transform yeast host cells and the positive transformed cells being isolated, followed by exffa,ction of the said second hybrid vector to obtain a sequence encoding a protein which binds to said intracellular domain binainst protein; and ('nc) a method for isolating and idettrifying a protein capable of binding to the intracellular domains of 'INF-Rs or F AS-R comprising applying the procedure of non stringent southern hybridization followed by PCR cloning, in which a sequence or parts thereof according to the invention is used as a probe to bind sequences from a cDNA or genomic DN.A library, having at least partial homolo~r thereto, said bound sequences then amplified and cloned by the PCR procedure to yield clones encoding proteins having at least partial homology to said sequences according to the invention.

4 PC'f/ITS95/05854 The present invention also provides a pharmaceutical composition for the modulation of the TNF- or FAS ligand- effect on cells comprising, as active ingredient, any one of the following (i) a protein according to the invention, or the protein p55IC, p55DD. FAS-IC
or FAS-DD, its biologically active fraemcnts, analogs, derivatives or mixtures thereof; (ii) a recombinant animal virus vector encodinE a viral surface protein capable of bindinb to~ a 'I i~'F-R or FAS-R - carrying cell - or tumor cell-specific receptor and a sequence encoding a protein.
analog or derivative of the invention or encoding the pS~IC, p55DD, FAS-IC or FAS-DD; (iii) a recombinant animal virus vector encoding a viral surface protein as in (ii) above and an oligonucleotide sequence encoding an antisense sequence of the p55IC, p55b17, FAS-1C or FAS-DD
sequence: and (iv) a l G vector encoding a ribozyne of sequence capable of interacting with a mR.'vA sequence encoding a protein, analog or derivative of the invention or a mRNA sequence encoding the p55IC, p55DD.
FAS-IC or FAS-DD.
A specific embodiment of the above aspects of the invention is the use of the p5~-IC or DN.A encoding therefor. This embodiment is based on the discovcr5~ that the p55-IC may in a 1 > ligand {'INF)-independem fashion induce other T1W'-associated effects in cells. Accordingly, there is provided a method for inducing TNF-associated effects in cells or tissues comprisins treating said cells with one or more proteins, analogs or derivatives thereof said one or more proteins being selected from a protein being essentially all of the sclf associatinJ
intracellular domain of the p55 TNF-R (p55-IC) or portions thereof capable of self associating and induang, in a ligand 20 ~-independent manner, said TNF effect in the cells, wherein said treating of the cells comprises introducing into said cells said one or more proteins, analogs or deriwtives in a form suitable for imracellular introduction thereof, or introducing into said cells a DNA sequence encoding said one or more proteins, analogs or derivatives in the form of a suitable vectar carrying said sequence, said vector being capable of effecting the insertion of said sequence into 25 said cells in a way that said sequence is expressed in said cells, Embodiments of the above method of the invention include {i) a method wherein said treating of cells is by transfection of said cells with a recombinant animal virus vector comprising the steps of (a) constructing a recombinant animal virus vector carrying a sequence encoding a 30 vial surface protein (ligand) that is capable of binding to a specific cell surface receptor on the surface of said cells to be treated, and a second sequence encoding a grotein being the p55-IC, portions thereof, analogs and derivatives of all of the foregoing, said protein when expressed in said cells being capable of self=association and induction of said one or more TIv'F-associated effects; and 35 (b) infecting said cells with the vector of (a).
Vii) a method wherein said x:NF effect to be induced in said cells is the induction of IL~8 gene e~.~pression, said.vector carrying a sequence encoding essentially all of said p55-1C, portions thereo>~ analogs and derivatives of all of the foregoing, which are capable, when expressed in the cehs of self association and sigrtaIing for the induction of said IL-8 gene expression.

(iii) a method for treating tumor cells or virally-infected cells, or for augmenting the antibacterial effect of granulocytes, wherein said viral vector carries a sequence encoding a viral Iigand capable of binding a specific cell surface receptor on the surface of said tumor cells, virallv-infected cells or pranulocytes and a sequence encoding said p5~-IC. portions thereof, analogs and derivatives thereof, which when expressed in said tumor, virally-infected or granulocytc cells induces TNF-associated effects leading to the death of these cells.
(iv) a method for treating tumor cells, wherein said p55-IC, portions thereof, analogs or derivatives thereof, when expressed in the tumor cells, induce the expression of IL-8 which leads to the killing of said tumor cells by its chemotactic activit~~ which attracts oranulocytes and other lymphocytes to the tumor cells resulting in the death of the tumor cells.
In this aspect of the invention, there is thus also provided the intracellular domain of the p55-R (p55-IC), portions. analogs and derivatives of all of the aforegoing for use in the treatment of cells by induction therein of TNF-associated effects; and the following embodiments thereof (i) the p55-IC, portions, analogs and derivatives for use in the treatment of cells by 1, induction therein of IL-8 Gene expression.
(ii) the p55-IC, portions, analogs and derivatives for use in the treatment of tumor cells by induction therein of IL-8 gene expression resulting in the killing of the tumor cells.
Moreover. in this aspeet of the invention there is provided a pharmaceutical composition for treating cells by induction therein of TNF-associated effects, comprising, as active in~,~redient, p55-IC, portions thereof analogs and derivatives of all of the aforegoing, and a pharmaceutically acceptable carrier; and the following embodimerns thereof (i) a pharmaceutical composition for treating cells by induction therein of TI~TF-associated effects, comprising, as active ingredient a recombinant animal virus vector encoding p55-IC, portions thereof, analogs and derivatives of all of the aforegoing. and a protein capable of binding a cell surface protein on the cells to be treated. _ ., _ .
(ii) a pharmaceutical composition for the treatment of tumor cells.
administration of said composition leading to the induction of IL-8 expression, and subsequent killing of the tumor cells.
As yet another aspect, the present invention provides a soluble, vligomeric tumor necrosis factor receptor (TIv'F-R) comprising at least two self-associated fusion proteins, each fusion protein having (a) at its one end, a TNF binding domain selected from the extracellular domain of a TNF-R, analogs or derivatives thereof, said extracellular domain, analogs or derivatives thereof being incapable of deleterious self association and being able to bind TIVF;
and (b) at its other gad, a self associating domain selected from (i) essentially all of the intracellular domain of the p55 ThIF-R (p55-IC), extending from about amino acid residue 206 to about amino acid residue 3~ 426 of the native pS~ TNF-R molecule (p55-R); (ii) the death domain of the p55-IC extending from about amino acid residue 328 to about amino acid residue 425 of the native p55-R{iii) essentially all of the intracellular domain of the Fas/A,F01 receptor (Fas-IC); (iv) the death domain of Fas-1C; and (v) analogs, fractions or derivatives of any one of (i)-(iv) being capable of self associaiiott, wherein said at least rivo self associated proteins self associate only at said ends ~ (b) having said ends (a) capable of binding to at least two TIFF monomers, each end (a) capable of binding one TNF monomer; and salts and functional derivatives of. said soluble, olis~omeric TNF-R.
Embodiments of this aspect of the invention include all of the above combinations of ends (a) with ends(b) ac defined above, for example, a soluble, oligomeric TI~IF-R
comprising as cxtracellular domain, the p55-R e~.-tracellular domain and as self=associating intracellular domain, the p55-IC.
Mareov cr, there is also provided a process for producing the soluble oligomeric flv'F-R of the invention comprising (a) the construction of an expression v ector encoding any one of said fusion proteins, the I0 DNA sequence of each of said ends of the fusion protein being obtained from cloned DNA
sequences encoding essentially all of said e~-tracellular domain of the T-~1F-R, analogs or derivatives thereof and from cloned DNA sequences encoding essentially all of said p~ 5-IC, p55-1C death domain, Fas-1C, Fas-IC death domain, analogs or derivatives of aII of the aforegoing, said ends beinc: Iigated together to form a fusion protein sequence, and said fusion protein sequence being inserted into said vector under the comrol .of transcriptional and translationai regulatory sequences;
(bj invoduction of the vector of (a) into a suitable host cell in which said fusion protein-is expressed; and (c) purification of the fusion protein expressed in said host cells, said fusion protein self associating prior to, during, or following the purification process to Sheld a soluble, oligomeric TNF-R
Furthermore, there is also provided a vector encoding the above fusion proteins, useful in the above method of the invention; host ctlls containing the vector, as «~cll as a pharmaceutical composition comprising the soluble, oligomeric 'T:~F-R, salts or functional dernatives thereof and mia-tures of any of the aforegoing according to the invention, as active in~~redient, together wiuu a pharmaceutically acceptable carrier. Similarly, the soluble, oligomeric 'fNF-R, salts, funetiorial derivatives thereof and mixtures of any of the aforegoing, according to the invention, are provided for use in antagonizing the deleterious effect of TNF in mammals, especiall~~
in the treatment of conditions wherein an excess of Ti'TF is formed endogenously or is e~ogenously administered; ar alternatively, for use in maintaining prolonged beneficial effects of T:~1F in mammals when used with TNF exogenously administered.
Along the Iines set forth concerning the above aspect of the invention, it has also been discovered that it is possible to construct a soluble, oIigomeric FasIAP01 receptor (Fas-R) which is useful for antagonizing the deleterious effects of the Fas ligand.
Accordingly, in a further aspect, the present invention provides a soluble, oligomeric Fas/:AI'O1 receptor (Fas-R) comprising at least two self associated fusion groteins, each fusion protein having (a) at its -one end, a Fas ligand binding domain selected from the txtracellular domain of a Fas-R analogs or dem~atives thereof being incapable of self associating and bcirie able to bind Fas ligand; and (b) at its other end,~a self associating domain selected from (i) esserttially all of the intracellular domain of the p55 ?NF-R (p55-IC), extending from about amino acid.rcsidue 206 to about amino acid residue 426 of the native p~~ TNF-R molecule (p55-R); (ii) the death domain of the p~5-IG
extending from about amino acid residue 328 to about amino acid residue ,26 of the native p35-R; (iii) essentially all of the intracellular domain of the Fas/AP01 r eceptor (Fas-IC); (iv) the death domain of Fas-IC; and (v) analogs or derivatives of any one of (i)-(iv) being capable of self association, wherein said at least two self associated proteins only self associate at said ends (b) having said ends (a) capable of binding to at least two Fas ligand monomers, each end (a) capable of binding vne Fas ligand monomer; and salts and functional derivatives of said soluble, oligomeric Fas-R.
In accordance with this aspect of the invention, there is also provided a process for the production of the soluble, oligomeric Fas-R comprising (a) the construction of an expression vector encoding any one of said fusion proteins, the DNA sequence of each of said ends of the fusion protein being obtained from cloned DNA
sequences encoding essentially all of said exuacellular domain of the Fas-R.
analogs or derivatives thereof , and from cloned DNA sequences encoding essentially all of said p5~-IC, p55-IC death domain, Fas-IC, Fas-IC death domain, analogs or derivatives thereof of all the aforegoin~, said ends being ligated together to for~rn a fusion protein sequence, and said fusion protein sequence being inserted into said vector under the control of transcriptional and translational regulatory sequences;
(b) introduction of the vector of (a) into a suitable host cell in which said fusion protein is expressed; and (c) purification of the fusion protein expressed in the host cells, said fusion protein self associating prior to, during, or following the purification process to meld a soluble, oligomeric Fas-R.
Moreover, also provided are an expression vector containing the fusion protein sequence encoding the soluble oligomeric Fas-R, useful in the above process; host cells ~.ontaining the vector; and pharmaceutical compositions comprising the soluble, oligomeric Fas-R salts or functional derivatives thereof or mixtures of any of the aforegoins as active ingredient together with a pharmaceutically acceptable carrier. Similarly, there is provided a soluble, oli?omeric Fas-R, salts or functional derivatives thereof or mixtures of any of the aforegoing, for use in antasonizing the deleterious e!~'ect of Fas ligand in mammals, especially in the treatment of conditions wherein an excess of the Fas ligand is formed endogenously or is exogcnously administered.
In a similar fashion to that noted above concerning the oligomeric TNF-Rs and oligomeric FAS-Rs, it is also possible to prepare mixed oligomers having binding specificity for both TNF
and FAS-R ligand. Thus, the present invention also provides a mixed oligomeric TNF-RIFAS-R
comprising at least two self associated fusion proteins, one of which fusion proteins is selected from any one of the above mentioned fNF-specific fusion proteins; and the other fusion protein is selected from any one of the above mentioned FAS-R Iigand-specific fusion proteins.. to provide a mixed oligomer having at least one TNF-R extracellular domain and at least one F.AS-R
extracellular domain associated by vrtuo of the self association between the intracellular domains ."
or portions thereof fused to each of these eYUacellular domains. These mixed olic:omeric receptors are prepared by preparing, as noted above, the oligomcric TNF-Rs and the oligomeric FAS-Rs and then mixing these totether and subsequently selecting, by standard procedures, those oligomers having binding specificity for both FAS-R ligand and TIv'F. Another way for preparing S the mixed oligomeric receptors is by co-transfecting suitable host cells «~th vectors, as noted above, encodinb any of the T1V'F-specific fusion proteins (soluble Tlv'F-Rs) and encoding any of the FAS-R ligand-specific fusion proteins (soluble FAS-Rs), purifying the expressed fusion proteins which self associate prior to, during. or following the purification to yield oligomeric receptors, and then selecting by standard procedures, those oligomeric receptors which arc 10 capable of binding to both TNF and FAS-R ligand.
Likev~~ise. there is also provided pharmaceutical compositions comprising the mixed oligomeric receptors, salts or functional derivatives thereof or mixtures of aw of the aforegoine as active in~~rediem together with a pharmaceutically acceptable carrier. In addition, there is provided the mixed oligomeric receptors, salts or functional derivatives thereof or mixtures of any 15 of the aforegoing, for use in antagonizing the deleterious effects of both T11F and F AS-R Iigand in mammals, especiahy in the treatment of conditions wherein an excess of TNF and FAS-R ligand is formed endogenvusl~~ or is exoLenously administered; or alternatively, for use in maintaining prolonged (slow-release) beneficial effects of TNF andlor FAS-R ligand in mammals when used with TNF and/or FAS-R Iigand (in soluble form) exogenously administered.
Other aspects and embodiments of the present invention are also provided as arising from the following detailed description of the invention.
It should be noted that, where used throughout, the following terms :
"Modulation of the TIv'F-effect on cells" and "Modulation of the FAS-ligand effect on cells" are understood to encompass in virrv as well as tn vivo treatment.
Brief Description of the brawings Figs la-c depict schematically the partial and preliminary nucleotide sequence of cDNA clones encoding the p55IC and p?SIC-binding proteins, wherein Fig. 1(a) is the sequence of clone 55.11 encoding the p55IC-binding protein 5.11; Fig. 7 (b) is the partial and preliminary sequence of clone ?5.3 encoding the p75IC-binding protein 75.3;
and Fig. 1(c) is the partial and preliminary sequence of clone ?5.16 encoding the p?5)iC-binding protein p?5.16; all as described in Example 1; and Fig. 1(d) depicts the deduced amino acid sequence of protein 55.11, deduced from the nucleotide sequence of Fig. 1(a), as also described in Example 1.
Fig. 2 is a reproduction of a Northern blot which shows the SS.I 1-specific mRNAs present in a number of tested cell lines, as described in Example 1.
Figs. 3A and B are reproductions of autoradiograms depicting the in vitro binding of the protein encoded for by the 55.11 cDNA to GST fusion proteins containing portions of p55-IC, wherein in Fig. 3A there is depicted the binding of the full-length 55.11 protein (55.11 full) to the various GST fusion proteins; arid in Fig. 38 there is depicted the bindinb of a portion of ~ 5.11 fused to the FLAG octapeptide to the various GST fusion proteins, all as described in Example 1.
Fig. 4 shows schematically a comparison of the deduced amino acid sequence of human 5.11 to related protein sequences derived from lower organisms, as described in Example 1.
S Fig. 5 is a reproduction of a Western blot stained with anti-MBP polyelonal antiserum, showing the self association of the pSSIC, the Western blot derived from an SDS-PAGE
gel on which were electrophoresed the interacting bacterially-produced chimeric proteins p55IC-MBP and p55IC-GST (lanes 1-4) or the control interaction between the chimeric protein pSSIC-MBP and GST alone (lanes ~-8), the interactions between the chimeric proteins (and control) being carried out on glutathion-agarose beads prior to SDS-PAGE, as described in Example 2.
Fig. 6 is a reproduction of phase contrast micrographs sho~zng the cyrtotoxic effect of the full-length pSSIC in HTtal cells transfected with an expression vector encoding this p~SIC
(right panel); and the inhibition of this cy~totoxic efrect when expression of the vector is blocked by treating the cells with tetracycline (left panel), as described in Example 2.
Fig. 7 depicts the ligand-independent triggering of the cvtocidal effect in HeLa cells transfected with the full-lens,~th p55-R, its intracellular domain, or pans of the intracellular domain including the 'death domain' where (i) at the extreme left hand side of Fig. 7 there is depicted schematically the various DNA molecules encoding the full-length pS5-R, its intracellular domain and the portions of the intracellular domain which were inserted into the vector with cvhich the HeLa cells were transfected.
(ii) the left and middle bar graphs show the TVF receptor expression in the HeLa cells of each of the types of receptor shown at the extreme left of Fig. ', the left bar representing the amounts of receptor in ng/ccll sample and the mid~Ie bar ~~~~~h representing the amounts of receptor c~cpressed in terms of radioiodinated TNF
bound to the transfected cells; and (iii) the right bat graph showing the viability of the HeLa cells expressing the various kinds of the receptor;
and wherein in all of the bar graphs the open bars represent cells transfected in the presence of tetracycline and the closed bars represent cells transfected in the absence of tetracycline; all of the above being described herein in Example 2.
Fig. 8 depicts the ligand-independent induction of IL-8 gene expression in HeLa cells transfected with the full-length p55-R or its intracellular domain (p55IC), wherein in panel A there is shown a reproduction of a Northern blot representing the Northern analysis of RNA
extracted from HeLa cells treated or untreated with TNF (tvvo left hand lanes marked 'control' and "ThIF'), and of RNA extracted from HeLa cells transfeeted with vectors encoding the p55-R, p55-IC or the control protein, luciferase (the remaining lanes marked 'p55-IC', 'p55 R' and Luc, respectively), the cells having been transfected in the presence (+) or absence (-) of tetracycline in each case (hence two lanes per transfection); and wherein in panel B there is shown the methylene blue stainin' of 18S rRNA in each of the HeLa cell sample shown in panel A; all of the abavc being described in Example 2.
Fig. 9 (A and B) depicts graphically the lieand independent triggering of a c;rtoeidal effect in HeLa cells transfected with pSSR or parts thereof, or u,~ith FAS-IC, u~hcrein in Fig. 9A
there is depicted the results with respect to the p55R or parts thereof and in Fig. 9B there is depicted the results with respect to the FAS-IC. In the left hand panels of both Fig. 9A
and B there is depicted schematically the portion of the p~SR or FAS-IC used in the transfections ~~hile the right hand panels depict graphically the experimental results, all as described in Example 2.
Fig. 10 depicts schematically the partial and preliminary nucleotide sequence of a cDNA clone, called T2', which encodes a protein capable of binding to the p55IC and p'AS-IC, as described in Example 3.
Fig. 11 depicts schematically the partial and preliminary nucleotide sequence of a cDNA clone, called F9, which encodes a protein capable of binding to the p»IC and FAS-IC, as described in Example 3.
Fig. 12 depicts sehematicallv the partial and preliminary nucleotide sequence of a cDl~'A clone, called DD11, which encodes a protein capable of binding to the p~~IC, especially the p55DD, and FAS-IC, as described in Example 3.
Detailed Description of the Invention The presem invention relates, in one aspect, to novel proteins vwhich are capable of binding to the intracellular domain of receptors belonging to the ?NF/NGF superfamily, such as TNF-Rs and FAS-R and hence are considered as mediators or modulators of this superfamily of receptors, e.g. of the TNF Rs and FAS-R, having a role in, for example, the signaling process that is initiated by the binding of TNF to the TNF-R and FAS Iigand to FAS R. Examples of these proteins are those which bind to the intracellular domain of the p55 TNF-R (p55IC), such as the proteins designated herein as 55.1: 55.3 and SS.11 (Example 1) as well as those encoded by cDNA clones F2, F9, and DDII (E.xample 3); those which bind to the intracellular domain of the p75 TNF-R
(p75IC), such as the proteins designated herein as 75.3 and 75.16 (Example 1 ); and those which bind to the intracellular domain of FAS-R (FAS-IC), such as the proteins encoded by cDNA
clones F2, F9 and DD11 (Example 3). Proteins 55.1 and ~~.3 have been found to represent portions or fragments of the intracellular domain of the p55 TNF-R (p55IC);
other proteins, 55.11, 75.3 and 75.16, represent proteins not described at all prior to the present invention (75.3, 75.1G) or those that have been described (55.11, see Khan et aL, 1992) but whose function and other characteristics, particularly, the ability to bind to a TNF-R, were not described in any way (see Example I, below). The new proteins encoded by cDNA clones F2, F9 and DD11 also represent proteins pre~.iously not described at all, i.e. their sequence is not in the'GENEBANK' or 'PROTEIN BANK' data banks of DNA or amino acid sequences.
Thus, the present invention concerns the DNA sequences encoding these proteins and the protans encoded by these sequences.

Moreover, the present invention also concerns the DhA sequences encoding biolo~eally active analogs and derivatives of these proteins, and the analogs and dcrivativcs encoded thereby.
The preparation of such analogs and derivatives is by standard procedure (see for example, Sambrook ct al., 1989) in which in the DNA sequences encodins these proteins, one or more codons may be deleted, added or substituted by another, to yield analogs having at least a one amino acid residue change with respect to the native protein. Acceptable analogs are those which retain at least the capability of binding to the intracellular domain of the T1V'FINGF receptor superfamily, such as FAS-R or TNF-R, e.g. the p55IC, p75IC or FAS-IC, or which can mediate any other binding or enzymatic activit3~, c.g. analogs which bind the p55, p75IC or FAS-IC but t 0 which do not signal, i e. do not bind to a further dowstream receptor, protein or other factor, or do not catalyze a signal-dependent reaction. In such a way analogs can be produced v~~hich have a so-called dominant-negative effect, namely, an analog which is defective either in binding to the, for example, p~SIC, p?SIC or FAS-IC, or in subsequent signaling following such bindinL. Such analogs can be used, for example, to inhibit the TNF- or FAS-ligand- effect by competing with the natural 1C-binding proteins. Likewise, so-called dominant-positive analogs may be produced which would sen~e to enhance, for example, the TNF or FAS ligand effect. These would have the same or better IC-binding properties and the same or better signaling properties of the natural IC-binding proteins. Similarly, deri~~atives may be prepared by standard modifications of the side S~raups of one or more amino acid residues of the proteins, or by conjugation of the proteins to . another molecule e.g. an anubody, enzyne, receptor, etc., as arc well known in the art.
The new' ?NF-R and FAS-R intracellular domain - binding proteins, e.g, the proteins SS.7 , 55.3, 5.11, 75.3, 75.16 as well as the proteins encoded by cDNA clones F2, F9 and DD11 (hercinafrer, F2, F9 and DD11) have a number of possible uses, for example.
(i) They may be used to mimic or enhance the function of T:~3F or F AS-R
Iigand, in 2~ situations where an enhanced 'T1~TF or FAS-R ligand effe.et is desired such as :n a~~ti-te.~;or.
anti-inflammatory or anti-FiIV applications where the TNF-or FAS-R ligand-induced cytotoxicity is desired. In this case the proteins, e.g. those binding to the p55IC such as S.I, 55.3, as well as F?, F9 and DD11, and the free p55IC itself (see below and Example 2), as v~~ell as the 'death domain' of the p55IC (p55DD), which enhance the TNF effect; or proteins F2, F9 and DD 11 as well as FAS-IC and F AS-DD which enhance the FAS-R
Iigand effect, i.e. cytoto~c effect, may be introduced to the cells by standard procedures I:now ger se. For example, as the proteins are intracellular and it is desired that they be introduced only into the cells where the TNF or FAS-R ligand effect is wanted, a rystem for specific introduction of these proteins into the cells is necessary. Onc way of doing this 3~ is by creating a recombinant animal virus e.g. one derived froth Vaccinia, to the D1~A of which the following two genes will be introduced the gene encoding a Iigand that binds to cell surface proteias specifically expressed by the cells e.g. ones such as the Alas (HIS j virus gp120 protein which binds specifically to some cells (CD4 lymphocytes and related leuketnias) or any other Iigand that binds specifically to cells carrying a TI'S-R or FAS-R, such that the recombinant virus vector wit! be capable of binding such T'NF-R-or FAS-R-WO 95/31544 PCT/US95l05854 carrlrin~ cells; and the gene encoding the new intracellula- domain-binding protein or the p55IC, p55DD, FAS-IC ar FAS-DD protein. Thus, expression of the cell-surface-binding protein on the surface of the virus will target the virus specincally to the tumor cell or other TNF-R- or FAS-R- carry~inc; cell, following which the intracellular domain-binding S protein encoding sequence or p55IC, pSSDD, FAS-IC or FAS-DD encoding sequence will be introduced into the cells via the virus, and once expressed in the cells will result in enhancement of the TNF or FAS-R Iigand ctfect leading to the death of the tumor cells or other TT~'F-R- or FAS-R- carrying cells it is desired to kill. Construction of such recombinant animal ~~ru5 iS by standard procedures (see for example, Sambrook et al., IO 1989). Another possibility is to introduce the sequences of the new proteins or the pS;IC, p55DD, FAS-iC or FAS-DD in the form of oligonueleotides which can be absorbed by the cells and expressed therein.
(ii) They may be used to inhibit the '1'VF or FAS-R ligand effect, e.g. in cases such as tissue damage in septic shock, graft-vs.-host rejection, or acute hepatitis, in which case it 15 is desired to block the TNF-induced TNF-R or FAS-R ligand induced FAS-R
intracellular signaling. In this situation it is possible, for example, to introduce into the cells, by standard procedures, oligonucleotides having the anti-sense coding sequence far these new proteins, or the anti-sense coding sequence for p55IC, p~~DD, FAS-IC ar FAS-DTI, which would effectively block the translation of mRN As encoding these proteins and 24 , thereby block their expression and lead to the inhibition of the T~F- or FAS-R ligand-eff'ect.
Such olisonucleotidcs may be introduced into the cells using the above recombinant virus approach, the second sequence carried by the virus being the oligonucleotide sequence. Another possibility is to use antibodies specific for these 25 proteins to inhibit their intracellular signaling activity. It is possible that these new~
proteins have an c~-tracellular domain as well as an intracellular one, the latter which binds to the TNF-R or FAS-R binding domain, and thus antibodies ?enerated to . their cxtracellular domains can be used to block their TNF- or FAS-R ligand- related functions.
Yet another way of inhibiting the TNF or FAS-R iigand effect is by the recently 30 developed ribozvme approach. Ribozymes are catalytic R~'~TA molecules that specifically cleave ItNAs. Ribozymes may be engineered to cleave target RNAs of choice, e.g. the m~t.NAs encoding the new proteins of the im~ention or the mRNA encoding the pSSIC, p55DD, FAS-IC or FAS-DD. Such ribazymcs would have a sequence specific for the mRUlA of choice and would be capable of interacting therewith (complementary bindinb) 35 folloR~ed by cleavage of the mRZVA, resulting in a decrease (or complete loss) in the expression; of the protein it is desired to inhibit, the level of decreased ea~pression being dependent upon the level of ribozyme expression in the target cell. To introduce ribozymes into the cells of choice (e.g. those carrying TNF-Rs or FAS-R) gay suitable vector may be used, e.g. plasmid, animal virus (retrovirus) vectors, that are usually used 40 for this purpose (see also (i) above, where the virus has, as second sequence, a cDl~A

2u ' encoding the ribozyme sequence of choice), Moreover, ribozwnes can be constructed which have multiple targets (mufti-target riboz;~mes) that can be used, for example, to inhibit the expression of one or more of the proteins of the invention and/or the p~SIC, psSDD, FAS-IC or FAS-DD as well (For reviews. methods etc. concerning ribozymes see Chen et al., 1992; Zhao and Pick, 1993; Shore et al., 199;. Joseph and Burke, 1993;
Shimayama et al., 1993; Cantor of al.., 1993; Barinaga. 199: ; Crisell et al., 1993 and Koimmi et al., 1993).
iii) They may be used to isolate. identify and clone other proteins which are capable of binding to them, e.g. other proteins involved in the intracellular signaling process that are downstream of the TNF-R or FAS-R intracellular domain In this situation, these options, namely, the DNA sequences encoding them may be used in the yeast two-hybrid system (see Example 1, below) in which the sequence of these protein: will be used as "baits" to isolate, clone and identify from cDNA or genomic DNA libraries other sequences ("preys") encoding proteins which can bind to these neH~ Tlv'F-R or FAS-R
intracellular domaitrbinding proteins. In the same way, it may also be determined whether the specific proteins of the present invention, nameE~~, those which bind to the p55IC.
p75IC, or FAS-IC, can bind to other receptors of the 'I~FIIvIGF superfamiU of receptors, For example, it has recently been reported (Schwalb et al., 7993; Baens et al.. 1993; Crowe et al., 1994) that there exist other T~1F-Rs besides the p55 and p75 'INTF-Rs. Accordingly, using the yeast two-hybrid system it may be specifically tested whether the proteins of the present invention are capable of specifically binding to these other TIv'F-Rs or other receptors of the ?NF/NGF superfamily. Moreover, this approach may also be taken to determine whether the proteins of the present invention are capable of binding to other known receptors in whose activity they may have a functional role.
ZS (iv) The new proteins may also be used to isolate, identif~~ and clone other proteins of the same class i.e. those binding to TIVF-R or FAS-R intracellular aomains or to functionally related receptors, and involved in the intracellular signaling process. In this application the above noted yeast two-hybrid system may be used, or there may be used a recently developed (fir ilks et al" I 989) system employing non-stringent southern hybridization followed by PCR cloning. In the Vililks et al, publication, there is described the identification and cloning of two putative protein-tyrosine kinases by application of non-stringent southern hybridization followed by cloning by PCR based on. the lozown sequence of the ldnase motif, a conceived kinase sequence. This approach may be used, in accordance with the preseni im~ention using the sequences of the new proteins to identify and clone those of related Ti'iF-R, FAS-R or related receptor (TNFI'NGF
superfamily receptors) intracellular domain-binding proteins.
(v) Yet another approach to utilising the new proteins of the invention is to use them in methods of affinity chromatography to isolate and identify other proteins or factors to which they arc capable of binding, e.e. other receptors related to TNF-Rs (TNF/NGF
receptor superfamily) or other proteins or factors involved in the intracellular signaling process. In this application, the proteins of the present invention, may be individually attached to afftnin~ chromatography matrices and thec: brought into contact with cell extract: or isolated proteins or factors suspected of being involved in the imracellutar signaling process. Following the affinity chromatography procedure, the other proteins or factors whaeh bind to the new proteins of the invention, can be eluted, isolated and characterized.
(«) As noted above, the new proteins of the invention may also be used as immuno~ens (antis:ens) to produce specific antibodies thereto. These antibodies may also be used for the purposes of purification of the new proteins either from cell extracts or from transformed cell lines producing them. Further, these antibodies may be used -for diagnostic purposes for identih~ing disorders related to abnormal functioning of the TNF
or FAS-R Iigand system; e. e. overactive or underactivc TNF- or F ~rS-R ligand-induced cellular effects. Thus, should such disorders be related to a malfunctioning intracellular signaling system involving the new proteins, such antibodies w outd serve as an important 1 S diagnostic tool.
It should also be noted that the isolation, identification and characterization of the nev~
proteins of the invention may be performed using any of the well known standard screeiiiii~
procedures. For example, one of these screening procedures, the yeast two-hybrid procedure asf is set forth in the following examples (Eacamples 1 and 3), was used to identify the new proteins of the invention. Likewise as noted above and below, ocher procedures may be employed such- as amity chromatography, DNA hybridization procedures, etc. as are well 1,-sown in the art, to isolate, identify and characterize the new proteins of the iwention or to isolate, identif~~ and characterize additional proteins, factors, receptors, etc, which are capable of binding to the aew~
proteins of the invention or to the receptors belongdng to the TIv'FlNCF
family of rc:eptors.=a si~~
As regards the antibodies mentioned herein throughout, the term "antibody" is meant 'to include polyclonal antibodies, monoclonal antibodies (mAbs), chimeric antibodies, anti-idiat5~pic (anti-Id) antibodies to antibodies that can be labeled in soluble or bound form, as well as fragnents thereof provided by any known technique, such as, but not limited to enzymatic cleavage, peptide synthesis or recombinant techniques.
Polyctvnal antibodies are heterogeneous populations of antt'bady molecules derived-from the sera of animals immunized with an antigen. ~. monoclonal antibody contains a substantially homogeneous gopulaiion of antibodies specific to antigens, which populations contains substantially similar epitope binding sifts. i~4Abs may be obtained by methods known to those skilled in the art. See; for exarnplc Kohler and Milstein, Nature, 2~G:495-497 (1975); U.S: Patent No. 4,376,114; Ausubel et al., gds., I3arIow and Lane ANTI$ODIES ' A
LABORATORY
MANUAL, Cold Spring Harbor Laboratory (1988); and Colligan et al., gds.;
Current Protocols in Immunology, Greene publishing Assoc, and Wiley Intcrscience N.Y., (1992, 1993).
Such antibodies may be of any immunollobulin class including fgG. IgM. IgE, IgA, GILD and any subclass thereof rA

hybridoma producing a m.~b of the present invention tnay be cultivated i»
arrn, i» siru or r» viva.
Production of high titers o~ mAbs in viva or in siru makes this the presently preferred method of pro.luction.
Chimeric antibodies are tnolecule.s different portions of which are derived from dif~r'erem animal species, such as those having the variable region derived from a marine tnAb and a human immunoglobulin constant repon. Chimeric antibodies are primarily used to reduce immunogenicity in application and to increase yields in production. for example, where marine mAbs have hislta yields from hybridomas but higher immunogenicity in humans, such that humanlmurine chimeric tttAbs are used. Chimeric antibodies and methods far their production are ' 10 known in the art (Cabilly et al., Pruc. Natl. Acctd .Sci. t :SA 81:3273-3277 (1984}; Motrison et al., Proc. h'arl. Acid ,Sci. LJ.~A 81:6851-685, (1984); Boulianne et al., ~'1~'axrrrc 312:643-(i46 (1984);
Cabilly et al., European Patent Application 125023 (published :vovember 14, 1984); Neuberger et al., ~'Vaturc~ 314:26S-270 ( 1985): Taniguchi et al., European Patent Application 171496 (published February 19, 1985); Marrisan et aL, European Patent Application 173494 (published March ~, 1980); ~ieuberger et al., PCT Application WO 8601533, (published March 13, 198G); Kudo et al., European Patent Application 184187 (published June 11, 1986); Sahagan 4t al., J Immunal.
13 i;lOG6-107: (1986); Rabinson et al., International Patent Application Ire.

(published May 7, 198'7); Liu et al., Prac. Natl. Acid Sci tlS.l 84:3439 X443 (1987); Sun et al., Proc. Natl. Acac~ Sci L~.4 84:214-218 (1987); Better et al., .Scienrx 240:1041-1043 (1988); and Harlow and Lane, ANTIBODIES :A LABORATOR~C' MANUAL, supra.
An anti-idiotypic (anti Id) antibody is an antibody which recoc_ntizes unique determinants generally associated with the antigen-binding site of an antibody. An Id antibody can be prepared by immunizing an animal of the same species and genetic type (e.g. mouse strain) as the source of the mAb with the mAb to which an anti-Id is being prepared. The immunized animal ' u,~il1 recognize and respond to the idiotypic determinants of the irnrnunizinl;
antibody by producing an antibody to these idiotypic deteaninants (the anti-Id antibody). See, for example, Lt.S. Patent No.
4,699, 880.
The anti-Id antibody may also be used as an "immunogen" to induce an immune respon~e in yet another animal, producing a so-called anti-anti-Id amibody. The anti-anti-Id may : be epitopically identical to the original mAb which induced the anti-Id. Thus, by using antibodies to the idiotypic deterntiaartts of a mAb, it is possible to identify other clones expressing antibodies bf identical specificity. , Accordingly, mAbs generated against the IC-bindinz proteins, analogs or dcriyativcs thereof of the present invention or the p55IC, p55DD, FAS-IC, FAS-DD, analogs or.derivatives thereof may be used to induce anti-Id antibodies in suitable animals, such as B ALB/c mice: Spleein cells from such iatcnunized mice are used to produce anti-Id hybridomas secreting ami-Id ~iii~bs.
Fuzther, the anti Id m.4bs can be coupled to a carrier such as keJfiole limpet hemocyanin {IQ.H) aid used to immunize additional BALBIc mitt. Sera from these mice will contain anti-anti'=Id antibodies that have the binding properties of the original mAb specific for an epitope ,of ;.the about IC-bindin_ proteins, analogs or derivatives yr p»IC, p:SDD, FAS-IC or F.4S-DD_ analogs or derivatives.
The anti-Id mAbs thus have their ow-n idiot5~pic epitopes, or "idiotopes"
structurall~~
similar to the epitope being evaluated, such as GRB protein-oc.
The term "antibody" is also meant to include both intact molecules a5 well as fragments thereof, such as, for example, Fab and Flab'), which are capable of binding antigen. Fab and F(ab'r fragments lack the Fc fragment of intact antibody, clear more rapidly from the circulation, and may have less non-specific tissue binding than an intact antibody (V~rahl et al., .I. Nucl. :'I~IcW
24:31b-325 (198;}).
It will be appreciated that Fab and F(ab')' and other fragments of the antibodies useful in the present invention may be used for the detection and quantitation of the IC-binding proteins or p55IC, p55DD. FAS-IC or FAS-DD according to the methods disclosed herein for intact antibody molecules. Such fra_c:ments are typically produced b~~ proteol~nic cleavage, using enzymes such as papain (to produce Fab fragments] or pepsin (to produce F(air')~ fragments}.
l~ An antibody is said to be ''capable of binding" a molecule if it is capable of specifically reacting with the molecule to thereby bind the molecule to the antibody, The term "epitope" is meant to refer to that portion of any molecule capable of being bound by an antihody which can also be rccoenized by that antibody-. Epitopes or "antigenic determinants'' usually consist of chemically active surface groupings of molecules such as amino acids or sugar side chains and have specific three dimensional structural characteristics as well as specific charge characteristics.
An "antigen" is a molecule or a portion of a molecule capable of being bound by an antibody which is additionally capable of inducing an animal to produce antibody capable of binding to an epitope of that antigen. An antigen may have one ar more than one epitope. The specific reaction referred to above is meant to indicate that the antiben will react, in a highly selcctivc manner, with its corresponding antibody and not with the multitude of other antibcicties which may be evoked by other antigens.
The antibodies. includin' fragments of antibodies, useful in the present invention may be used to quantitatively or qualitatively detect the IC-binding proteins or p55IC, p55DD, Fr~S-IC, FAS-DD in a sample or to detect presence of cells which express the IC-binding proteins of the preseat invention or the p55IC, p55DD, FrIS-IC, FAS-DD proteins. This can be accomplished b;~
immunofluorescence techniques employing a fluorescently labeled antibody (see below) coupled with light microscopic, flow cytomctric, or fluorometric detection.
The artibodies (or fragments thereof) useful in the present invention may be employed histologically, as in inllnunofluorescence or immunoelectron microscopy, for irt situ detection of 1C-binding proteins of the present invention or the pSSIC, p~3DD, FAS-IC. FAS-DD. hl sine detection may be accomplished by remo~~ino a histological specimen fram a patient, and providing the labeled antibody of the present invention to such a specimen. The antibody (or fragment) is preferably provided by applying or by overlaying the labeled antibody (or fragment} ~'to a biological sample. Through the use of such a procedure, it is possible to determine not only the presence of the IC-binding proteins or the p55IC, p55DD, FAS-IC, FAS-DD, but also its distribution on the e~:amined tissue. Using the present invention, taose of ordinay skill will readily perceive that any of wide variety of histological methods (such 3s staining procedures) can be modified in order to achieve such in situ detection.
Such assays for IC-binding proteins of the present im~ention or the p~~IC, p55DD, FAS
IC, FAS-DD, typically compzises incubating a biological sample. such as a biological fluid, a tissue extract. freshly harvested cells such as hmphocrtes or leukocytes, or ells which have been incubated in tissue culture, in the presence of a delectably labeled antibody capably of identifying the IC-binding proteins or the p55IC, p55DD, FAS-IC. FAS-DD, and detecting the antibody by any of a number of techniques well knov~~n in the art.
IO = The biolopcal sample may be treated with a solid phase support or carrier such as nitrocellulose, or other solid support or carrier which is capable of immobilizing cells, cell particles or soluble proteins. The support or carrier ma~~ then be washed v~ith suitable buffers followed by treatment with a detestably labeled antibody in accordance with the present invention, as noted above. The solid phase support or carrier may then be washed vvith the buffer a second I ~ time to remove unbound antibody. The amount of bound iabel on said solid support or carrier may then be detected by conventional means.
By "solid phase support", "solid phase carrier", "solid support", "solid carrier", "support"
or "carrier" is intended any support or carrier capable of binding antigen or antibodies. Well-lmown supports or carriers, include glass, polystyrene, polypropylene, pol;~ethylene, deatran, 20 nylon amylases. natural and modified celluloses, golyacrvlamides, gabbros and magnetite. The nature of the carrier can be either soluble to some ea-tent or insoluble for the purposes of the present imrention. The support material may have virtually any possible structural configuration so long as the coupled molecule is capable of binding to an antigen or antibody.
Thus, the support or carrier configuration may be spherical, as in a bead, cylindrical, as in the inside surface of a test 25 tube, or the external surface of a rod. Alternatively, the surface may be flat such as a sheet, test strip, etc. Prcfetted supports or carriers include polystyrene beads. Those skilled in the art will know may other suitable carriers for binding antibody or antigen, or will be able to ascertain the same by use of routine experimentation.
The bindir~ activity of a gives lot of antibody, of the invention as noted above, nosy be 30 determined according to well known methods. Those skilled in the art will be able to determine operative and optimal assay conditions for each determination by employing routine experimentation_ Other such steps as washing, stirring, shaking, filtering and the like may be added to the assays as a customary or necessary for the particular situation.
35 One of the ways in which an antibody in accordance with the present invention can be deteaably labeled is by Iinldng the same to an enzyme and use in an enzyme immunoassay (E1A).
This enzyme, in tuna, whoa laxer exposed to an appropriate substrate, will react with the substrate in such a manner as to produce a chemical moiety which can be detected, for example, by spcctrophotometrie, fluorometric or by ~risual means. Enzymes which can be used detectably.label 40 the antibody include, but are not limited to, tnalate dehydrogenase, staphylococcal nuclease, delta-5-steroid isomeras, yeas alcohol dehydrogenase, alpha~~ly,:erophosphate dehydroeenase. triose ' phosphate. isomerase, horseradish peroudase, alkaline phosphatase, asparaginase, glucose oxidase, beta-galactosidase, ribonuclease, unease, catalase, glucose-6-phosphate dehydrogenase, glucoamylase and acetylcholinesterase. The detection car be accomplished by colorimctric methods which employ a chromogenic substrata for the enzyme. Detection may also be accomplished by visual comparison of the extent of enzymatic reaction of a substrate in comparison with similarly prepared standards.
Detection may be accomplished using any of a ~~arierrr of other immunoassays.
For example, by radioactivity labeling the antibodies or ancibod~~ fra_~ments, it is possible to detect R
PTPase through the use of a radioimmunoassay (RIA). A good description of RIA
may be found in Laboraton Techniques and Biochemistry in Molecular Hiology, by Work, T.S.
et al., Notch Iiolland Publishing Company. NY (1975) with particular reference to the chapter entitled "An Introduction to Radioimmune Assay and Related Techniques" by Ghard, 'T.
The radioacti~ a isotope can be detected by such means ~as the use of a y counter or a scintillation counter or by autoradio~~raphy.
It is also possible io label an anribodv in accorda~:ce with the present invention with a fluorescent compound. When the fluorescently labeled antioody is exposed to light of the proper wavelength, its presence can be then detected due to fluorescence. Among the most commonly used fluorescent labeling compounds are fluorescein isothiocyanate, rhodamine, phycoerythrinc, pycocyanin, altophycocyanin, o-phthaldehyde and fluorescamine.
The antibody can also be detestably labeled using fluorescence emitting metals such gas 1528, or others of the lamhanide series. These metals can be attached to the antibody using such.
metal chelating groups as diethyienetriamine pentaacetic acid (ETPA).
The antibody can also be detestably labeled by coupling it to a chemiluminescent compound. The presence of the chetniluminoscont-tagged antibody . is then deterrriiried ~Fk~y detecting the presence of luminescence that arises durine the course of a chemical reaction.
Examples of particularly useful chemih~minescent Labeling compounds are iuminol, isoluminol, thcromatic accidinium ester, imidazolt, acridinium salt and o~:alate ester.
Likewise, a bioluminescent compound may be used to label the antibody of the present imremion. Bioluminescence is a type of chemiluminescence found in biological systems in-v~ihich''a catalytic protein increases the efficiency of the chemiluminescent reaction.
The presence_'of a bioluminescent protein is determined by detecting the presence of luminescence. Important bioluminescent compounds for purposes of labeling are luciferin, luciferase and aequorin.
An antibody molecule of the present invention may be adapted for utilization yin 'an immunometzic assa~,~, also Imow-n~as a "two-site" or "sandwich" assay. in a t3~pical immunometric assay, a quantity of unlabeled antibody (or fragment of antibody) is bound to a solid ~ support ~or carrier and a Quantity of detestably labeled soluble antibody is added to permit detection"aridtor quantitation of the ternary complex formed between, solid-phase antibody, antigen, and labeled antibody.
* Trade-mark WO 95/31544 °CT/IIS95/05854 .~.b Typical, and preferred, immunometric assays include "forward" assays in which the antibody bound to the solid phase is first contacted with the sample being tested to exuact the antigen from the sample by formation of a binary solid phzse antibody-antigen complex. After a suitable incubation period, the solid support or carrier is v~ashed to remove the residue of the fluid sample, including unrcacted antigen, if any, and the contacted with the solution containing an unknown quantity of labeled antibody (which functions as a "reporter molecule"). After a second incubation period to permit the labeled antibody to comply with the antinen bound to the solid support or carrier throueh the unlabeled antibody, the solic support or carrier is washed a second time to remove the unreacted labeled antibody.
In another type of "sandwich" assay, which may also be useful with the antigens of the present invention, the so-called "simultaneous" and "ret~erse" assays are used. A sirnuItaneous assay involves a single incubation step as the antibody bound to the solid support or cattier and labeled antibody are both added to the sample being tested at the same time.
After the incubation is completed, the solid support or carrier is washed to remove the residue of fluid sample and 1 S uncomplexed labeled antibody. The presence of labeled antibody associated with the solid supper, or cagier is then determined as it would be in a conventional "fonvard"
sandwich assay.
In the "reverse" assay, stepwise addition nrst of a solution of labeled antibody to the fluid sample followed by the addition of unlabeled antibody bound to a solid support or carrier after a suitable incubation period is utilized. After a second incubation, the solid phase is washed in conventional fashion to free it of the residue of the sample being tested and the solution of unrcacted labeled amibody. The determination of labeled antibody associated with a solid support or cattier is then determined as in the "simultaneous" and "forward" assays.
The new proteins of the invention once isolated, identified and characterized by any of the standard screening procedures, for example, the ~~cast two-hybrid method, affinity chromatography, and any other well known method l:nov~m in the art; may then bP produced :hy any standard recombinant DNA procedure (see for example, Sambrooh, et al., 1989) in which suitable eukaryotic or prokaryotic host cells arc transformed by appropriate eukaryotic or prokaryotic vectors containing the sequences encoding for the proteins.
Accordingly, the prcsem invention also concerns such expression vectors and transformed hosts for the production of the proteins of the invention. As mentioned above, these proteins also include their biolos~cally active analogs and derivatives, and thus the vectors encoding them also include vectors encoding analogs of these proteins, and the transformed hosts include those producing such analogs. The derivatives of these proteins arc the derivatives produced by standard modification of the proteins or their analogs, produced by the transformed hosts.
In another aspect, the invention relates to the use of the free intracellular domain of the p55 TNR-R (pSSIC) or FAS-R (FAS-IC) or their so-called 'death domains' (p~SDD
or FAS DD, respectively) as an agent for enhancing the TNF or FAS-R ligand effect on cells, on its ow-n .(see Example 2). Where it is desired to introduce a ThTF- or FAS-R-ligand- induced cytotoxic effect in calls, e.g. cancer cells or HIV-infected cells, the p»IC, p55DD, FAS-IC or FAS-DD can be introduced into such cells using the above noted (see (i) above) recombinant animal vine (e.g.

vaccinia) approach. Mere too, the native p55IC. pS~DD. r AS-IC or FAS-DD, biololricaiiy active analogs and derivatives or framnents may be used, all of which can be prepared as noted above.
Likewise, the present invention also relates to the specific blocking of the TNF-effect or FAS-R ligand-effect by blocking the activity of the p55IC, pSSDD, FAS-IC ar FAS-DD, e.b. anti s sense oligonucleatides may be introduced intb the cells to block the expression of the pSSIC, p55DD, FAS-IC or FAS-DD.
The present invention also relates to pharmaceutical compositions comprising recombinant animal virus vectors encoding the TNF-R ar FAS-R intracellular domain binding proteins (,including the pSSIC, p55DD, FAS-IC and F.4S-DD), which vector also encodes a virus surtace 1Q protein capable of bindinb specinc target cell (e.c. cancer cells) surface proteins to direct the insertion of the inuacellular domain binding protein sequences into the cells.
In another aspect, the present invention also concerns, specifically, the effects of the self associating intracellular domain of the p~5 TNF receptor i g~5-IC, see Example 2~. An example of such effects, which is an effect normally mediated by T\'F binding to its receptor and which is 15 mimicked by the signaling activity of the self associating p~5-IC or parts thereof, is the induction of expression of the gene encoding IL-8.
h.-8 is a cytokine belonging to the subclass of chemokiaes having primarily cheinotactic , activity, and has been shown to play a major role in the chemotaxis of ~,~ranulocytes and other cell types associated with a number of patholocdcal states (see for example, Endo et al., 199; Sckido 20 et aL, 1993; Harada et al., 1993; Fcrrick et al., 1991).
TNF has a beneficial activity, and is used as such, in treatments to destroy tumor ells and virus i,~ected cells or to augment antibacterial activities of granulocytes.
However, as noted above, TNF also has undesirable activities in which case it is desired to block its activity, inch~ding those situations where large doses of TNF are used in cancer therapy, antiviral, therapy 25 or antibacterial therapy.
Accordingly, it is desirable to be able to direct TVF or a substance capable of mimicking iu beneficial activity to the cells ar tissues that it is specifically desired to treat.
In accordance with the present iwention it has been found that the self associating intracellular domain of the p55-R (p55-IC) can, in a ligand-independent manner, mimic a number 30 of effects ofTNF, e.g. the 'death domain' of p55-IC can induce cytotoxic effects an cells, and that the p55-IC can induce IL-8 gene expression. Thus, it is possible to utilize the p55-IC to mimic TNF function in a site-directed fashion, i.c. to introduce the p55-IC only to those cells ar tissues it is desired to treat.
One example of the above approach, as mentioned above, is to specifically transfect.
3~ (transform) tumor cells ar malignant tissue with a DNA molecule encoding p~~-IG or a portion thereof which can induce not only cytotoxic effects on such cells or tissue but also augment these effects by the co- -induction of IL-8, which W 11 result in the accumulation at the site of these cells or tissue of s~anulacytes and other lymphocytes, which, in turn, will serve to destroy the tumor cells or tissue. This approach obviates the need for administration of large doses of TNF with its 40 as.~cociated deleterious side-effects.

WO 95/31544 . , PCT/US95/05$54 as Using conventional recombinant DNA technoloc,~. it is possible to prepare various regions of the p55-IC and to determine which region is responsible for cacti TNF-induced effect, e.g. we have determined that the 'death domain' is responsihle far cytotoxicity (Example 2), and we have already prepared various other constructs containing f onions of the p55-IC, which portions (tobether with part or all of the death domain) may be responsible fox other TNF-effects, arid which may be used in a Iigand-independent manner, once self associated for activity, to induce these effects, e.g. IL-8 induction.
It should be noted that the sequeace of the p55-IC involved in the induction of other TNF
associated effects (e.g. IL-8 induction) may be different to that involved in cytotoxicity, i.e, may inctude none or only part of the 'death domain' and hav a other sequence motifs from other regions of the intracellular domain, or may be the same sequence. different features of the sequence !same sequence motif) being involved in the induction of different effects.
Accordingly, as detailed above and below, expression vectors coataining these p~5-IC
portions, analogs or derivatives thereof may be prepared, expressed in hose cells, purified and tested for their activry. In this way, a number of such p~~-IC fragments having one or more TNF
associated activities may be prepared and used in a ditierential fashion for the treatment of any number of pathological conditions, e.g. viral infections, bacterial infections, tumors, etc. In all of these situations the speeihe activity can be augmented b~- incorporation (or co-transfection) with the p~5-IC fragment responsible for IL-8 gene expression induction, permitting the desirable IL-8 chcmotactic activity to enhance the destruction of the~cclls or tissues it is desired to destroy.
Thus, without adcriinistering systemically Th'F, it is possible to induce its desirable effects by specifically introducing all or part of the p55-IC into the cells or tissues it is desired to treat.
The p55-IC may be introduced specifically into the cells or tissues it is wished to destroy by any one of the abovementioned procedures. For example, one way of doing this is by creating a recombinant animal virus e.g. one derived from Vaccinia, to whose DNA the followi~ti ~~~
genes will be introduced ' the gene encoding a ligand that binds to cell surface proteins specifically expressed by the cells e.g. ones such as the AD7S virus gp120 protein which binds specifically to some cells (CD4 lymphocytes and related lcukemias) or any other ligand that binds specifically to cells carrying a TNF-R such that the recombinant virus vector will be capable of binding such TIvTF-R-carrying cells; and the gene encoding the p55-IC or a portion thereof. Thus, expression of the cell-surface-binding protein on the surface of the virus W
11 target the virus specifically to the tumor cell or other TNF-It-carrying cell, following which the p55-IC, or portion thereof, encoding sequence will be introduced into the cells via the virus, and once expressed in the cells will result in enhancement of the TNF effect leading to the death of the 3~ tumor cells or other 'ITrF-R-carrying cells it is desired to !:ill or induction, for example, of IL-8 which will lead to cell death. Construction of such recombinant animal virus is by standard procedures (see for example, Sambrook et al., 1989), Another possibility is to introduce the sequences of the p55-IC or parts thereof in the form of oligonucleotides which can be absorbed by the cells and expressed therein.

The present invention thus also relates speci5cally to pharmaceutical compositions comprising the above recombinant animal virus vectors encoding the p55-IC or portions thereof, which vector also encodes a virus surface protein capable of binding specific target cell (e.g.
cancer cells) surface proteins to direct the insertion of the p~5-IC, or portions thereof, sequence S into the cells.
The present invention relates, in yet another aspect, to new synthetic T11F
receptors which are soluble and capable of oligomerization to form dimerie, and possibly also high order multimeric, TNF receptor molecules, cacti monomeric part of these receptors being capable of binding to a TNF monomer. T~'F occurs naturally as a homotrimer containing three, active ?NF
monomers, each capable of binding to a single T1'F receptor molecule, while TNF receptors occur naturally as monomers each capable of binding only one of the monomers of the TNF
homotrimeric molecule. Thus, when TNF binds to TyF receptors on the cell surface, it is capable of binding to three receptor molecules resulting in the clustering of the TN'F
receptors, which is believed to be the start of the silrnaling process which ultimately triggers the observed TNF
effects on the cells.
While T'I'TF has many desirable effects such as its abiliy to destroy, for example, tumor cells or virus-infected cells and to augment antibacterial activities of granulocytes, TNF does however, have many undesirable effects such as, fur example, in many severe diseases including autounmune disorders, rheumatoid arthritis, graft-versus-host reaction (graft rejection), septic shock, Ti'1F has been implicated as the major cause for pathological tissue destruction. TNF may also cause excessive loss of weight (cachexia) by suppressing the activities of adipocytes.
T4oreover, even when administered for its desirable activities, e,g. in the treatment of various malignant or viral diseases, the dosages of TNF used arc often high enough to cause within the patient a number of undesirable cytotouc side-effects; e.g. the destruction of healthy tissue.
Accordingly, in all of the above instances where TNF action is undesirable, an effective inhibitor of TNF has been sought. bZarty TNF-blocking agents have been proposed, including soluble proteins capable of binding TNF and inhibiting its binding to its receptors and hence also inhibiting the cyZOtoxic effects of TNF (see EF' 3083'78, fiP 398327 and EP
568925). However, these TNF binding proteins, or soluble TNF receptors are monomeric, each binding only one of the TNF monomers of the TNF homotrimer. Hence, the blocking of the 'fl~'F
function may not be complete, each monomeric receptor-bound T1~IF' molecule still having two TGIF
monomers free to be able to bind cell-surface TNF receptors and illicit its effects on the cells.
In order to overcome the above drawbacks in blocldrtg TNF function, there has been developed in accordance with the present invention a means for constructing, as fusion proteins, soluble oligomeric TNF receptors which ate capable of binding at mast two TTW
monomers of the naturally occurnng TNF homotrimer molecule. As a consequence, these soluble oIigomeric 1NF
receptors bind more.avidly to their TNF Iigand than the previously known monomeric soluble TNF binding proteins or receptors. For example, v~~hen the soluble T;VF
receptor of the invention is in the form of a dimer, it is capable of binding two T:VF monomers of a TVF
trimer and hence do C~,~S°° ~ .»'.e ~.~rnnleya ,~a~trst:~~tt~~ CC rho 'T't~'_ ~i~ic ~~utr?'lIZ?ttOr! 11e"1~ mOrP ~llStP!nef~

WO 95/31544 . PCTlUS95/05854 because of a lower dissociation rate of the dimeri,: soluble receptors from the TNF. Moreover, such soluble, oligomeric receptors are also large: than their monomeric counterparts and thus, pharmaceutically, they are also advantageous because of the likelihood of their having a slower clearance rate from the body.
5 The basis for the development of the soluble oligomeric TIFF receptors of the invention, was the discovery that the intracellular domain of the p55-R TNF receptor was capable of self association, and further, that within this intracellular domain (p55-IC) there exists a region, the so-called 'death domain'. which is also capable of self association and as such, in a ligand-indepcndent fashion, can cause cytotoxic effects en cells (see Example 2) Utilizing this self 0 association properly of the p55-IC and its 'death domain' it is thus possible to construct a fusion protein. using standard recombinant DNA technolo~~y~, containing essentially alI of the e~ctracellutar domain of a TNF receptor such as the p75-R or p55-R receptors, preferably the p5~-R, and fused thereto, essentially all of the intracellular domain (p55-IC) or the death domain of the p5~-IC. Im this way a nea~ fusion product is produced which has ai one end the TNF binding S domain i.e., the extracehular domain of the receptor, and at its other end the intracellular domain or the death domain thereof which is capable of self association. Accordingly, such a product can oligomerize by self association between two (and possible more) p55-1C or death domains thereof to yield oligomers (or at least dimers) having at Ieast nvo TNF binding domains.
Furthermore, it has also been discovered in accordance with the present invention, that the ?0 FaslA.PO1 receptor has a self associating, intracellular domain inclusive of a self associating 'death domain' haring certain homolofy to the pSs-IC and death domain thereof (Example 2).
Accordingly, it is possible to construct the soluble, oligomeric TNF receptors of the invention by fusing the exuacellular domain of the TNT' receptor (as noted above) to the intracellular domain or the 'death domain' of the FasIAP01 receptor.
~5 In both of the above noted situations, the oligomaric TNF receptors of the invention are soluble by virtue of having only the soluble extcacellular domain of the TNF
receptor and the soluble intracellular domain or death domain .thereof of either the p55 R TN-F
receptor or the FasIAP01 receptor, i.e. they do not contain the transmembranal {insoluble) domain of either type of receptor.
;0 The construction of the about oligomcric TNF receptors of the invention are detailed heron below in Example 4. It should however be noted that upon construction of the oligomeric TNF receptors of the invention, there may arise a situation, heretofore not reported, that the extraceilular domain of the TIv'F receptor is capable of self-association, a situation that may not be desirable as it could interfere with the ability of the oligomeric receptor to bind to t~vo or more Th'F monomers of the TNF homotrimeric molecules or may lead to less than optimal binding ,of such TNF monomers. Accordingly, in such a situation, it is possible, by standard recombinant DNA procedures, to modify the eh-tracellular domain of the TNF receptor by, for example, . deleting or substituting one or more amino acid residues contained within the self associating repon to prevem such self association. Such modifications of the ea-cracellular domain of the TNF
receptor -are thus also part of the present invention and arc designated herein as analobs or WO 95131544 . PCTIUS95/05854 derivatives of the e~.-cracellular domain of the T_VF receptor. In a similar fashion, the self associating intracellular domain (IC) or death domain (DD) Thereof of the p55-R receptor or the FasIAP01 receptor used in the oligomeric TNF receptors o~the invention, may also be analogs or derivatives thereof i.e. may be any modification of the p55-IC sequence or portions thereof including the death domain (p55DD}. or any modification of the Fas/~PO1 intracellular domain (FAS-iCj sequeace or portions thereof including the death domain (FAS DD}, providing that these modifications yield a self associating product.
Similarly, once produced and purified, the soluble oligomeric TNF receptors, analogs or derivatives thereof. may be further modified b5~ standard chemical means to provide salts and functional derivatives thereof for the purposes of preparing pharmaceutical compositions containing as active ingredients these TNF receptors of the invention.
For the production of the soluble, oligomeric TiVF receptors of the invention, the DNA
sequences encoding the e~.-tracellular domain of the TNF receptor are.
obtained from existing clones of the entire TNF receptor, as is the intracellular domain or death domain thereof, and as is also the intracellular domain or death domain of the Fas,%APO1 receptor (see Example 2 and Example 5). In this u:ay the DNA sequence of the desired extracellular domain is ligated to the DNA sequence of the desired intracellular domain or portion thereof including the death domain, and this fused product is inserted (and Iigated) into a suitable expression vector under the control of the promoter and other expression control sequences. Once formed, the expression vector is introduced (transformation, transfeetion, etc.) into a suitable host cell, which then expresses the vector to yield the fusion product of the invention being the soluble self associating T~1F receptor molecules. These arc then purified from the host cells by standard procedures to yield the final product being the soluble, oligomeric TNF receptors.
The preferred preparation of the fusion product encoding the ea-tracellular domain and intracellular domain or portion thereof is by way of PCR technology using otigonucleorides specific for the desired sequences to be copied from the clones encoding the entire TNF receptor molecule. Other means are also possible, such as isolating the desired portions encoding the extracellular domain and the intracellular domain, by restriction endonucleascs and then splicing these together in a known fashion, with or W thout modifications at the terminal ends of the restriction frar_aments to ensure correct fusion o~ the desired portions of the receptor (ek-tracellular and intracellular domains or portions thereofj. The so-obtained fusion products are then inserted into the eoprcssion victor of choice. ~-In a similar fashion, the present invemion also concerns soluble, oligomeric Fas/AP01 (FAS) receptors containiag the e:ctracellular domain of the FasJAP01 receptor and the self associating intracellular domain of the p55-R (p55-IC), the death domain theceof (p55DD), or the self associating intracellular domain of the Fas/AP01 receptor (FAS-)iC) or the death domain thereof (FAS DD), or any analogs or derivatives thereof (see above}. The construction of these soluble, oligomeric FAS receptors is detailed in Example ~ herein below, using an available cloned full-length FAS receptor-encoding sequence as starting material and the appropriate oligonucleotides for PCR production of the desired extracellular and intracellular domains, follen:ed by ligation thereof to yield a fusion product. which is then inserted into a suitable ea-pression vector. ?~s detailed above and belov.~. prokaryotic or euharyotic vectors and host cell may be used to produce the desired soluble, oIieomeric FAS receptors, which can then be purified and formulated, as active ingredient. into a phartaaceutical composition.
S The above soluble, oligomeric FAS receptors of the invention are intended for effective blocking of the Fas lieand, which may also exist as a trimer (similar to T~fF, see above), each oligomeric receptor of the invention capable o: binding two or possible more Fas ligands and thereby neutralize their aeti«ty. The Fas ligand is 1.-sown to be predominantly cell-surface associated but may also exist in a soluble form. In any event, the oligomcric F AS receptors of the invention can bind to at least two monomers of this lictand and thereby neutralize more effectivel~~
(than monomeric FAS receptors) the activiry of the Fas Iigand. The Fas ligand, and hence activation thereby of the FAS receptor, has been implicated in a number of pathological states, particularly those relating to liver damage (apoptosis of hepatocwes, for example), includine liver damage associated with hepatitis, as well as ir, autoimmune conditions, including lymphocyte damage (apoptosis) in HIV-infected humans (see, for example Ogasawara et al., 1993, Cheng et al., 1990. Accordingly, the soluble, olieomeric F AS receptors of the invention are intended for blocking the activity of Fas ligand and ma~~ be used as active ingredient in pharmaceutical compositions for ueating such Fas ligand-associated patholotical states.
Likewise, the present invention also concerns soluble, oligomeric receptors which have binding affinity for both TNF and FAS-R Iigand, the so-called "mixed" TNF
RIFAS-R oligomeric receptors. These mixed oligomeric receptors will contain at least one Tl~'F-R
extracellular domain and at least one FAS-R extracellular domain which arc associated in the oligomeric recoptar by virtue of each of these exuaceIlular domains being fused t~ any one of the above-mentioned, self associating, p55IC, p55DD. FAS IC or FAS DD
These mixed oligomeric receptors may be prepared by : (a) providia? a-y of the a.'~ove noted fusion products which contain the exuacellular domain of a TNF-R (p75 TNF-R or preferably, p55 TNF-R) fused to any one of the self associating intracellular domains p55 IC and FAS IC or an}~ ono of the self associating 'death domain' p55DD and FAS DD, or any self associating portions, analogs or derivatives of any thereof; (b) providing any of the above noted fusion products which contain the extracellular domain of FAS-R fused to any one of the self associating p55IC, FAS-IC, p55DD, and FAS DD, or any self associating portions, analogs or derivatives of any thereof and (c) mixing any of the TNF-specific fusion products of (a) with any of the FAS-R ligand-specific fusion products of (b) to provide (following standard selection and purification procedures) oligomeric (dimeric or higher order oligomeric) receptors which have at least both the extracellular domains ef a T:VF-R and FAS-R that are associated by.virtue of the self association capability of their fused IC or DD regions.
Another possibility for the preparation of the above mixed oligomeric receptors is by co-transforming suitable host cells with the above-mentioned expression vectors, one of which encodes the TNF-specific TNF-R fusion products and one of which encodes the FAS-R ligand-specific FAS-R fusion products. Following the e~.~pression of these different fusion products in the host cells, the mixed oligomeric (TIFF-RT'AS-Ri receptors may be obtained by standard purification and selection procedures.
The utility of these mixed affinity oligome~c receptors is primarily for the neutralization of both T'!~1F and FAS-R Iigand when these are o~r:r-expressed endogenously or are at undesirably high levels folloHZnb exogenous adminisvation. Recent evidence points to a likelihood that there exists a synergism in function between the FAS-R ligand (usually cell-surface associated) and TNF-a (which may also be cell-surface associate;:). Accordingly, in some instances it is desired to neutralize both of these Iibands at the same point on the cell surface, i.e.
such a mixed-amity receptor can block both the TNF bindine to its r eceptor and the binding of FAS-R ligand to its receptor. Accordineiy, these mixed-afl7niy receptors may be used as an active ingredient in pharmaceutical compositions for treating such conditions (see above) where both TIFF and FAS-R Ii~and effects are undesirable.
Similarly, along the lines mentioned abo~~e concerning the soluble, oligomeric TNF-R and FAS-R, and mixed TNF-R/FAS-R olieomers or the invention, it is also possible to produce 1 ~ soluble, oiigomeric receptors for other receptor., or any mixtures thereof, in particular those of any of the other members of the TNFr'NGF super family. In this case, any of the extracellular domains of the various receptors can be fused to the above-mentioned self associating inuacellular domains or portions thereof or to anv other intracellular domains of the super family members also capable of self association.
fixpression of any of the recombinant proteins of the invention as mentioned herein can be effected in eukaryotic cells (e.g. yeast, insect or mammalian cells), using the appropriate expression vectors. Any method known in the arc may be employed.
For example, the DNA molecules coding for the proteins obtained by any of the above methods are inserted into appropriately constructed expression vectors by techniques well known in the art (see Sambrook et al., 1989). Double-stranded cDI~'A is linked to plasr!id vectors by homopoh~tneric tailing or by rcstzictioo linl.-ing involving the use of synthetic DNA linkers or blunt-ended ligation techniques. Dh:~ ligases are used to ligate the DIv'A
molecules and undesirable joining is avoided by treatment W th alkaline phosphatase.
In order to be capable of expressing the desired protein, an expression vector should comprise also specific nucleotide sequences containing transcriptional and translational regulatory information linked to tbc DNA. coding for the desired protein in such a way as to permit gene expression and production of the protein. First, in order for the gene to be transcribed, it must be preceded by a promoter recognizable by RNA polymerise, to which the polymerise binds and thus initiates the transcription process. There are a ~~ariety of such promoters in use, which work with different efficiencies (strong and weak promoters). They are di~'erent for prokaryotic and ;:;~;;~.
eukaryotic cells.
The promoters that can be used in the present invention may be either constitutive; for example the int promoter of bacteriophage 0, the ~1_a promoter of the ~3-lactamasc gone ~ of pBR322, and the CAT promoter of the chloramphenicol acetyl transferase gene of pPR325, etc., or inducible, such as the prokaryotic promoters including the major right and left promoters of WO 95/31544 Pt"T/US95/05854 bacteriophage ~ (PL and P~, the trp,, sec 4 la=?. 1~, ~, and ga_l promoters of E. cell, or the trpllac, hWrid promoter, etc, lGlick, B.R. ( 198 : . Besides the use of strong promoters to generate large quantities of mRNA, in order to achieve f~gh le4els of gene expression in prokaryotic cells, it is necessary to use also ribosome-binding sites to ensure that the mRNA is efficiently translated.
One example is the Shine-Dalgarno sequence (SD sequence) appropriately positioned from the initiation codon and complementary to the 3'-terminal sequence of iGS RNA.
For eukaryotic hosts, different transcriptional and transiational regulatoy sequences may be employed, depending on the nature of the host. They may be derived from ~~iral sources, such as adenovrus, bovine papilloma virus, Simian ~:rus, or the like, where the regulatory signals are associated with a particular gene which has a high level of expression.
Examples are the ?K
promoter of Herpes virus. the SV44 early promoter, the yeast gall eeae promoter, etc.
Transcriptional initiation regulacott~ si~ntals may be selected which allow for repression and activation. so that expression of the ~:encs can be modulated.
. The DNA molecule comprising the nucleotide sequence coding for the fusion product proteins of the invention is inserted into a vec:or having the opeiably linked transcriptional and translational regulaton~ signals which is capable ef integrating the desired gene sequences into the host cell. The cells which have been stable transformed by the introduced DNA
can be selected by also introducing one or more markers which allow for selection of host cells which contain the expression vector. The marker may pro~~ide for phototrophy to an auxotropic halt, biocide resistance, e.g. annbiotics, or hea~~~ metals, such as copper, or the like.
The selectable marker . gene can either be directly linked to the DNA gene sequences to be expressed, or introduced into the same cell by co-transfection. Additional elements may also be needed for optimal synthesis of proteins of the invention. These elements ma5~ include transcription promoters, enhancers, and termination si~.Tnals. cDNA expression vector s incorporating such elements include those described by Okayama, H. (1983).
In a preferred embodiment, the imroduced DNA molecule will be incorporated into a plasmid or viral vector capable of autonomous re;pIication in the recipient host. Factors of importance in selecting a particular plasmid or viral vector include : the ease with which recipient cells that contain the vector may be reco_anized and selected from those recipient cells which do not contain the vector; the number of copies of the vector which arc desired in a particular host;
and whethex it is desirable to be able to "shuttle" the vector between host cells of different species.
Preferred prokaryotic vectors include plasmids such as those capable of replication in $.
toll, for example, pBR322, CaIEI, pSC101, pACYC 184, etc. (see Maniatis et al., 1982;
Sambrook et al., 1989); Bacillus plasmids such as pC194, pC221, pT127, etc.
(Gryczan, T..
{1982)); Streptomyces plasmids including pIJ101 (Kendall, K.J. et al., (1987)); Streptomyces bacieriophages such as QtC31 (Chalet, K.F. et al., in : Sixtl~Intornational Syrrl,aosium on ~rc~otrrvcetales Biolosri. (1986)), and Pseudomonas plasmids (lohn, J.F. et al., (1986), and Izald, K. (1978)). Preft~rred eukaryotic plasmids include BPV, vaecinia, SV40, 2-micron circles, etc., or their derivatizes. Such plasmids are well known in the aft (Botstein, D. et al., (1982);

JJ
Broach, J.R. in : The Molecular Biology, of the feast Saccharomvc~s : Life Cvcle and Inheritatics (1981); Broach, J.R, (1982); Bolton. D.P. et al., (1980); Maniatis, T., in :
Celf Biolo~rv v A
Co~,rehensive Treatis~yol. 3 ' Gene Ex ression, (1980); and Sambrook et al., 1989).
Once the vector or DhIA sequence containing the constructs) has been prepared for expression, the D~1A construet(s) may be introduced into an appropriate host cell by any of a variety of suitable means : transformation. transfection, conjugation, protoplast fusion, electroporation, calcium phosphate-precipitation, direct microinjection, etc.
Host cells to be used in the invention may be either prokaryotic or eukaryotic. Preferred prokaryotic hosts include bacteria such as E. c;nli, Bacillus, Streptomyces, Pscudomonas, Salmonella, Serratia, etc. The most preferred prokaryotic host is E. ~uli.
Bacterial hosts of particular interest include E. evh K1? strain 294 (ATCC 314-46), E. cwli X1776 (A,TCC 31537), ,~ cull ~~'3110 (F-, lambda-, prototropic (_~TGC 27;'?5)), and other enterobacterium such as Salmonella typhimurium or Serratia marcescens and various Pseudomonas species.
Under such conditions, the protein will not be glycosylated. The prokaryotic host must be compatible with the replicon and control sequences in the expression plasmid.
Preferred euharyotic hosts are mammalian cells, e.g. human, monkey, mouse and Chinese hamster ovan~ (CHO) cells, because they provide post-translational modincations to protein molecules including correct folding or glycosylation at correct sites. Also yeasts cells can carry out post-translational peptide modifications including I;Iycosylation. ~
number of recomb'uiant DNA strategies adst which utilize strong promoter sequences and high copy number of plasmids which can be utilized for production of the desired proteins in yeast. Xeast recognizes leader sequences on cloned mammalian gene products and secretes peptides bearing leader sequences (i.e., pre-peptides). ' After the introduction of the vector, the host cells are grown in a selective medium, which selects for the growth of vector-containing cells. Expression of the cloned gene sequences) results in the production of the desired proteins.
Purification of the recombinant proteins is carried out by any ane of the methods known for this purpose, i.e. any convrntional procedure involving extraction, precipitation, chromatogaphy, electrophoresis, or the like. A further purification procedure that may be used in preference for purifying the protein of the invention is affinity chromatography using anti-TNI"
receptor monoclonal antibodies, which are produced and immobilized on a gel matrix contained within a column. Impure preparations containing the recombinant protein are passed through the column. The protein will be bound to the column by the specific antibody while the impurities will pass throufih. After washing, the protein is eluted from the gel by a change in p):I or ,ionic 3 5 strength.
As used herein (see above), the term 'salts' refers to both salts of carboxyl groups and to acid addition salts of amino groups of the protein molecule formed by means known in the art.
Salts of a carboa~yl group include inorganic salts, for example, sodium, calcium, and salts with organic bases as those formed, for example, ro~~ith amines, such as trieihanolamine, arginine or lysine. Acid addition salts include, for example, salts with mineral acids and salts with organic acids.
"Functional derivatives" as used herein covers derivatives which may be prepared from the functional groups which occur as side chains on the residues or the N- or C-terminal s~~roups, by S means known in the art, and are included in the invention as long as they remain pharmaceutically acceptable, i.c. they do not destroy the activist- of the protein and do not confer toxic properties on compositions containing it. These derivatives include aliphatic esters or amides of the carbox-hl groups, and N-acyl derivatives of free amino groups of 0-acyl derivatives of free hydroxyl groups formed with acyl moieties (e.g. alkanoyl or carboc~~clic aroyl groups).
10 -"Fractions" as used herein refers to am~ part or portion of the receptor.
(intracellular or extracellular domains thereof), or of the proteins binding to the intracellular domain of the receptor, provided it retains its biotol,~ical activity.
As mentioned above, the present invention also relates to various pharmaceutical compositions comprising a pharmaeeuticall~~ acceptable carrier and the various noted active 15 in~edients of the invention or their salts, functional derivatives, or mixtures of any of the foregoing. These compositions may be used in am' of the conditions as noted herein, for example, in conditions where there is an over production of endogenous TNT', such as in cases of septic shock, cachexia, daft-versus host reactions, autoimmune diseases like rheumatoid arthritis, etc.
The way of administration can be via any of the accepted modes of administration for similar 20 agents and will depend on the condition to be treated, e.g. when used to inhibit TNF effects they may be administered intravenously in case of septic shock or local injection in case of rhwmatoid arthritis (for example, into the knee), or continuously by infusion, etc. The compositions may also be used, for example, in cases of TNF intoxication caused by exogenous administration of excessive amount (overdoses) of TNF, e.g. in the case of cancer therapy or oral disease therapy.
25 The pharmaceutical compositions of the invention are prepared for adr~inistratie~ h:~
mixing the protein or its derivatives with physiologically acceptable carriers, stabilizers and excipients, and prepared in dosage form, e.g. by lyophilization in dosage vials. The amount of active compound to be administered will depend on the route of administration, the disease to be veatcd and the condition of the patient. For example, local injection in case of inflammatory 30 conditions of rhwm$toid arthritis wdl require less'activ~ ingredient on a body weight basis~than wt~l iatravcnous infusion in case of septic shock.
Other aspects of the invention will be apparent from the following examples.
The invention will now be described in more detail in the following non-limiting exacxiples and the accompanying drawings .,,, _.
CloninE and isolation of Qroteins which bind to the intracellular domains of the p55 and Q75 T1VF receptors To isolate proteins interacting with the intracellular domains of the p55 and p75 TNF
receptors (p~SIC and p75 IC), the yeast two-hybrid system was used (Fields and Song, 1989).

WO 95/31544 pCT/US95/05854 3"
Briefly, this two-hybrid system is a yeast-based genetic assay to detect specific protein-protein interactions i~: vivo by restoration of a eukaryotic transcriptional activator such as GAZA that has 1<vo separate domains, a DNA binding and an activation domain, which domains when expressed and bound together to form a restored Gr~t4 protein, is capable of binding to an upstream activating sequence which in turn activates a promoter that controls the expression of a reporter gene, such as lacZ or HIS3, the expression of which is readily observed in the cultured cells. In this rystem the genes for the candidate interacting proteins are cloned into separate expression vectors. In one expression vector the sequence of the one candidate protein is cloned in phase with the sequence of the G.4L4 DNA-binding domain to generate a hybrid protein with the GALA
DNA-binding domain, and in the other vector the sequence of the second candidate protein is cloned in phase with the sequence of the GAL4 activation domain to ?enerate a hS~brid protein with the GrlL4-activation domain. The two hybrid vectors are then co-transformed into a yeast host strain having a lacZ or HIS3 reporter gene under the control of upstream GAL4 binding sites. Only those transformed host cells (cots ansformants) in which the two hybrid proteins are expressed and are capable of interacting with each other. will be capable of expression of the reporter gene. In the case of the lacZ reporter gene, host cells e~cpressing this gene vviIl became blue in color when 7~-gaI is added to the cultures. Hence, blue colonies are indicative of the fact that the two cloned candidate proteins arc capable of interacting with each other.
Using this two-hybrid system, the intracellular domains p55IC and p75IC were cloned, . separately, imo the vector pGBT9 (carrying the GAL4 DNA-binding setaucncc, provided by CLONTECH, USA, see below), to create fusion proteins with the G4L4 DIVA-binding domain (similarly, the intracellular domain, FAS~IC and a portion of the 55IC, namely, the 55DD were also cloned into pGBT9 and used to isolate other IC-binding proteins, see Example 3 below). For the cloning of p55111; and p75IC into pGBT9, clones encoding the full-len~zth cDlrTA sequences of p55 TNF-R (Schall et al., 1990) and p75 TNF-R (Smith et al., 1990) were used from which the intracellular domains (IC) were excised as follows : p55IC was excised using the enzymes EcoRI
and SaII, the EcoRI-SaII fragment containing the p55IC sequence was then isolated by standard procedures and inserted into the pGBT9 vector opened, in its multiple cloning site region (IvICS), with EcoRI and SaII. p75 IC was excised using the enzymes BspHI and SalI, the BspFII-SaIT
franment containing the p75 IC sequence was then isolated by standard procedures arid filled-in with the klenow enzwte to generate a fragment which could be inserted imo the pGBT9 vector opened with Smal and SaII.
The above hybrid (chimeric) vectors were then cotransfected (separately, ~ one cotransfection with the p55IC hybrid and one with the p75 IC hybrid vector) together with a 33 eDNA library from human HeLa cells cloned into the pGAD GH vector, bearing the GAL4 activating, domain,, into the HF7c yeast host strain (all the above-noted vectors, pGBT9 and pGAD GH carrying the HeLa cell eDNA library, and the yeast strain were purchased 'from Clontech Laboratories, Inc., USA, as a part of MATCFaviAhER Two-Hybrid System.
#PT1265-1). The co-transfected yeasts were selected for their ability to grow in medium lacking Histidine (Ii'ts- medium), growing colonies being indicative of positive transformants.
The selected yeast ~s clones were then tested for their abilin~ to ex~rzss the lacZ gent, i.e. for their LAC Z activity, and this by adding X-gal to the culture medium. v~~hich is catabolized to form a blue colored product bl~ ~-galactosidase, the enzyme encoded by the lacZ gene. Thus, blue colonies are indicative of an active lacZ gene. For activity of the IacZ gene, it is necessary that the GAL4 transcription activator be present in an active form in the transformed clones, namely that the GAL4 DNA-binding domain encoded by one of the above h3~brid vectors be combined propcrlv with the GAL4 activation domain encoded by the other hybrid vector. Such a combination is only possible if the two proteins fused to each of the GALS domains are capable of stably interacting .(binding) to each other. Thus, the I-bs+ and blue øAC Z~ colonies that were isolated are colonies which have been cotransfected with a v actor encoding p~SIC and a vector encoding a protein product of human HeLa cell origin that is capable ofbindin~ stably to pas IC; or which have been transfccted with a vector encodinb p75IC and a vector encoding a protein product of human HeLa cell origin that is capable of binding stably to p75 IC.
The plasmid DNA from the above His", LAC Z" yeast colonies was isolated and 1 S eleccroporatcd into E. coG strainI~ 1 O 1 by standard procedures followed by selection of Leu+
and ~Ampicillin ~ resistant transformants, these transformants being the ones cam~ing the hybrid pGAD GH vector which has both the AmpR and Leu~ coding sequences. Such transformants therefore are clones carrying the sequences encoding newly identified proteins capable of binding to the p551C or p75IC. Plasmid DNA was then isolated from these transformed E.
coli and recested by (a) retransforming them with the original intracellular domain hybrid plasmids (hybrid pGTB9 carrying either the p55IC or p7~IC sequences) into yeast strain HF7 as set forth hereinabove. As controls, vectors carrying irrelevant protein encoding sequences., e.g. pACT-lamin or pGBT9 al5ne were used for corraasformation with the p55IC-binding protein or p75IC-binding protein encoding plasmids. The cotransformed yeasts were then tested for c.~owth on His-medium alone, or with different levels of 3-aminotriazole; and (b) recraasforming the plasmi.d DNA and original intracellular domain hybrid plasmids and control plasmids described in (a) into yeast host cells of strain SF'Y526 and determining the LAC Z'~ activity (effectivity of ~-gal formation, i.e. bhxe color formation).
the results of the above tests revealed that the pattern of graw-th of colonies, in His-medium was identical to the pattern of LAC Z actiy~, as assessed b; the color of the colony, i.e.
I3is+ colonies were also LAC Z+. Further, the LAC Z activity in liquid culture (preferred culture conditions) was assessed after transfection of the GAL4 DNA-binding and actin anon-domain hybrids into the SFX52G yeast hosts which have a better LAC Z inducibilit5r with the GAL4 transcription activator than that of the HF-7 yeast host cells.
The results of the above co-transfections are set forth in Table 1 below, from which it is apparent that a number of proteins were found that were capable of binding to the p55IC; or the p75IC, namely, the proteins designated 55.11, which binds to the p55IC; and 75.3 and .75.16 . which bind to the p75IC. All of these p55IC- and p75IC-binding proteins are authentic human proteins all encoded by cDNA sequences ori~~inating from the IicLa cell cDNA
library, which .

were fused to the G.aL,4 activation-domain sequence in the plasmid pG.AD GH in the above yeast two-hybrid analysis system Interestingly, it was also found that fragments of the p55IC, itself, namely, the proteins designated 55.1 and 55.3 were capable of binding to g55IC. These are discussed also in Example 2 below.
~.BLE 1 SUMMARY OF THE CIiARACTERISTICS OF SOME OF THE
cDNA CLONES (SEE ALSO EhAMPLE 3) ISOLATED Bl' THE
TWO-HYBRIDSl'ST~DZ APPROACH
DNA-binding Activation- Colon3~ colorLac Z Rcti<~itc~
; ~ in domain hybrid domain hybrid li uid culture assay GBT9-IC55 --- white ' 0.00 GBT9-iC55 55.1 blue 0.65 GBT9-IC55 _ _ blue 0.04 55.3 --- 5 S. I white 0.00 --- i 5 white 0.00 S.s ~

ACT-Lamzn _ white ' 0.00 _ 55.1 ACT-Lamin __ white 0.00 55.3 GBT9 55.1 white 0.00 !

GBT9 ~ 55.3 0.00 w ' white BT9-IC55 ( 55.11 ; blue ND

55.1 I white ND

ACT-Lamin SS.11 white ND

GBT9 55.11 white 1'TD

GBT9-IC75 75.3 blue Iv'D

GBT9-IC75 -- w-hitc ND

--- 75.3 ; white ND
i ACT-Lamin 75.3 white ND

GBT9 75.3 white ND

GBT9-IC7> 75.16 blue ND

75.16 white ND~

ACT-Lamin 75.16 white ND

GBT9 75.16 white ND

In the above Table 1, the plasmids and hybrid encoding the G.4L4 DNA-binding.dornain and GAL4 activation domain are as follows : ;
pNA-binding domain hybrids pGBT9-IC55 : full-length intracellular domain of the p55-TNF-R (p55IC) pACT-Lamin : irrelevant protein - lamin.
pG13T9 : vector alone pGBT9-IC75 : full-length intracellular domain of the p75-TNF-R (p75IC) . ~-~ ~
>~::~:aax Activation-domain hybrid : ... ;;~~a;--v~~~
55.1 and 55.3 correspond to fragments of the intracellular domain of the p55-TIv'F-R.
55.11 : is the novel protein associating with the p55-TIFF-R
75.3 and 75.16 are the novel proteins associating with the p75-TNF-R.

' The above noted cloned cDl~:~ encoding the novel p~~IC- and p75IC- binding proteins, 55.11, ?5.3 and 75.16, were then sequenced using standard DNA sequencing procedures. The partial sequence of all of these protein-encoding sequences is set forth in Figs. 1 a-c, where Fig.
1(a) depicts the sequence of the cDNA encoding protein 55.11; Fig. 1(b) depicts the partial S sequence of the eDNA encoding protein 75.3; and Fig. 1(c) depicts the partial sequence of the cDNA encoding protein 75.16. In Fig. 1(d) there is shown the deduced amino acid sequence of the protein 55.11, as deduced from the nucleotide sequence of Fig. 1(a).
It should be noted. however, that a partial sequence of the eDNA encoding the 55.11 protein has also been reported by Khan et al. (1992), in a study of human brain cDNA sequences, which study was directed at the establishment of a new rapid and accurate method for the sequencing and physical and genetic mapping ef human brain cDNAs. However, Khan et al. did not provide any information as regards the function or any other characteristics of the protein encoded by the 55.11 eDN A sequence; such functional or other analysis not being the intention of lvhan et aI. in their study.
1 s Analysis and characterization of the 5S 11 protean fal_ neral Procedures and Materials ail Ctonin~ of the cD~I4 of 55 11 Upon the analysis {for example, Northern Analysis -see below) of the cDNA of protein 55.1 I, it was revealed that the above noted 55.11 cDNA cloned by the two hybrid scrctn procedure represented only a partial cDNA of 35.11 having nucleotides 925-2863 (see Fig. 1(a)) which code for amino acids 349-900 (see Fig. 1(d)). The remaining part of the 55.11 cDNA (nucleotides 1-924 (Fig. 1(a)) which code for amino acids 1-308 (Fig. 1(d))] was obtained by standard procedures. namely, by cloning by PCR from a human fetal liver cDNA
library (for more details, see below). The full nucleotide sequence of 55.11 (Fig. 1(a)) was determined is both directions by the dideoxy chain termination method.
fiil Two-hybrid ~patactosidase expression tests ~i-galactosidase expression tests were performed as described above, except that in some of the tests, the plfl' I6 vector, which contains the activation domain of VP16, . was used instead of pGAD-GH, the Gal4 activation domain vector. Numbering of residues in the proteins encoded by the cDNA inserts are as in the Swiss-Prot*data bank.
Deletion mutants were produced by PCR, and point mutations by oligonucleotide-directed mutagenesis (Kunkel, 1994).
{iiil Northe n analysis Total RNA was isolated using TRI REAGENT (Molecular Research Crnter, Inc., Cincinnati, Oh., U.S.A.), denatured in formaldehyde<'formamide <-buffer, electrophoresed through an agaroselformaldehydc gel, and blotted to a GeneScreen.-..Plus*
membrane (Dupont, Wilminston, De., U.S.A.) in IOxSSPE buffer, using standard techniques. The blots were hybridized with the partial cDNA of 55.11 (see above, nucleotides 92~-28G3), radiolabeled with .the random-prime kit (Boehringer Mannheim Biochcmica, Mannheim, Germany), and washed stringently. Autoradiography was performed for I week.
* Trade-mark a~
jivl Ez~re~sion of 55.11 eD'~ 4 in HeLa cells and binding of the 55.11 protein toil to hione S-transferstse ~us~nroteins of p55'IC
Glutathione S-transferase (GST) fusions with p55-IC (GST-p55IC) and with p55-IC truncated below amino acid 345 (GST-p55IC:45) were produced and adsorbed to glutathione-agarose beads as described in Example ? below (see also Smith and Corcoran, 1994;
Frangioni and Neel, 1993). The cDNAs of 55.11 (1-2863 nucleotides, i.e. the full-length SS.i1 cl7NA), of FLAG-55.11, and of fuciferase were expressed in HeLa cells. FLAG-55.11 is the region extending; between residues 309 and 900 in the 55.11 protein (the partial cDNA of 55.11 (nucleotides 925-28b3), originally cloned by the two hybrid screen), N-linked to the FLAG
octapeptide (Eastman Kodak, New Haven, Ct., U.S.A.). -Expression of the fusion proteins was accomplished using a tetracycline-controlled expression vector (HtTA-1 ) in a HeLa cell clone that expresses a tetracycline-controhed transa:.tivator (see Example 2 below, and Gossen and Bujard, 1992). Metabolic labeling of the expressed proteins with j35S] T2et and [3oSj Cys (Dupont, Wilmington, De., U.S.A, and Amersham, Buckinghamshire, England), lysis of the HeLa cells, immunoprecipitation, and binding of the labeled proteins to the GST fusion proteins were performed as described below (Example 2), except that 0.5°~a rather than 0.1% Nonidet F-40 was present in the cell lysis buffer. The immunoprecipitations of 55.11 and FLAG-55.11 were achieved using a rabbit antiserum (diluted 1:500) raised against a GST fusion protein containing the region of 55.11 that ekrtends between amino acids 309 and 900 and a mouse monoclonal antibody against the FLAG octapeptide (M2; Eastman Kodak; 5 ~eglml of cell Iysate).
fbl Binding of the 55.11p tro ein to~55-I~aithin transformed veasts In this study it w-as sought to ascertain the nature of the binding between 5.11 and p55IC, in particular, the re~dons of both of these proteins involved in this binding. For this purpose the above two-hybrid procedure was used in which various full-length and deletion 2S mutants of p55IC (see also Example 2 below) in "DNA-binding domain"
constructs were used as "baits" to bind the "preys", being the partial SS,I1 protein encoded in constructs in which the partial 55.11 sequence (residues 309-900, as originally isolated) was fused to the "activation domain" in the vectors GAL4AD and VP16AD. Further, various deletion mutants of 55.11 were also constructed and fused to the "activation domain" in the GAL4Al7 vector (e.g. mutants of 55.11 having only residues 309-6S0 and 457-900). The binding of the various 'binding domain' constructs to the various 'activation domain' constivcts was examined in transferred SFY526 yeast cells. The binding was assessed by a two-hybrid ~-galactosidase expression filter assay. The non-relevant proteins SNFI and SNF4 screed as positive controls for the 'binding domain' and 'activation domain' constructs, respectively; the empty Gal4 (pGAD-GH) and VP
1 G (pVP 1 G) vectors sen~ed as nebative controls for the 'activation domain' constructs;
znd the. empty Gal4 (pGBT9) vector served as a nebative control for the bindinb, domain' constructs. The results of the assay are set forth in Table 2 below in which the symbols "~" and "J-:-"
indicate the development of strong color within 20-60 min of initiation of the assay, respectively (positive binding results); and "-" indicates no development of color within 24h of commencement of the assay (negative results). Blank spaces in the Table indicate binding assays not tested.
* Trade-mark WO 95131544 4 2 pCf/US95105854 T le 2 Binding of the 55.11 protein to p55-IC
within transformed yeasts D
o m s~d~1 I
z o' m o r + + +
a ~ °~' + + + ' t a >

z . ~~,o q t Q ~ ~y .
> m ors ~ +
V m °0 a z sos~ t a r °89,6 t °°s'sp + + + + + t t t t + +
.~
_~
U ~ M c° M ~ ~ Ii , 9 ~ O O O 'd t~~D N
N N N N N M
2~ ~' r .
(_~ ~_ zm d~~ .
.~
O~, ;~ a ;~.;
z o . ~,~~
o ~a~
- ,~02 .
...

From the results presented in Table 2 above it may be concluded that ».11 binds to p55-IC at a site which is distinct from the 'death domain' (residues 328-426) of p.SS-IC.
The 55. i 1 protein bound to a truncated p5~-IC from which the death domain had been deleted (construct 206-328 in 'Table 2). more effectively than to nontruncated p55-IC. It also bound to an even further C terminally truncated construct (construct 206-308j and to a construct from which both the death domain and a membrane proximal part were deleted (construct 243-328), However, the 55.11 protein did not bind to a construct that was N-terminally truncated down to amino acid 266 (Table 2). These findings indicate that the binding site for 55.11 is located in the regrion that extends between residues 243 and 308 of p55-IC and that the N
terminus of this binding site is between residues 243 and 266.
Transfer of the cDNA for 5.11 from the originally cloned 'prey' construct, which contained the Gal4 activation domain, to a prey construct containing the VP 16 activation domain did not decrease the binding efbcienc~~ of the 55.11 protein to p55-IC (Table 2). Thus, the structures) involved in tlus binding appear to reside within the 55.11 molecut and not to involve the site of fusion of 55.11 with the acti~~ation domain However, binding of 55.11 to p55-IC was abolished by even limited truncations of the 55.11 protein at either its C (55.11 construct 309-680) or N terminus (55.11 construct 457-900).
(residue 309 is the first residue in the 55.1I protein encoded by the partial cDNA clone oril,~nally isolated in the two hybrid saecn).
The observed binding between 55.11 and p55-IC appeared to be specific since SS.I 1 did not bind to other proteins, including three receptors of the Th~'INGF receptor family (p75-R, FaslAPO1 and CD40) and other proteins such as lamin and cyclin D (data not shown). It should be noted that of the other TIv'F/NGF receptor proteins tested there was also tested portions thereof which include their intracellular domains : human FAS-R (residues 175-319), CD40 (residues 216-277) and p75-TNF-R (residues 287-461), none of which bound 55.11 (data not shown).
~cl Northern analysis of the RN4 frarn several call Lines, using the 55.11 cDN4 ac a probe and cloninE of the full-legit ~ 55.1LcDNA
The cell lines examined were HeLa, CEM, Jurkat, and HepG2 cells derived from human epithelial carcinoma, as acute lymphoblastic T cell leukemia, an acute T
cell leukemia, and a hepatocellular carcinoma, respectively. The 55.11 cDNA original isolated (nucleotides 9? 5 2863) was used as a probe. Samples consisted of 10 ~S of RNA/lane. The results of the Northern analysis are shown in Fig. 2, which is a reproduction of a Northern blot.
From Fig. 2 it is thus apparent that the Northern analysis using the 55.11 cDNA as a probe revealed, in several cell lines, a single hybridizing transcript of about 3 kB, which is larger than the cDNA (2 kB) of the originally isolated 55.11 cDNA. Using oligonucleotide primers that correspond to the 55.11 sequence, we cloned by PCR a 5' extending sequence whose len~,nh was about 1 kB. The sum of the length of this 5' e~ctcnding sequence W th that of the originally cloned cDNA approximates the length of the 55.11 transcript. The 3 1;B cDNA that encompassed both WO 95131544 PCT/fJS95105854 these portions w~as effectively expressed in transfected HeLa cells (see belou~) yielding a protein of about s4 kDa, which suggests that the 3 1~B cDN~, contains a translational start site.
d In vitro bindino f the 55.11 rotein to GST-fusion roteins containin onions of 5 S To ascertain that 55.11 can indeed bind to p~5-IC and to exclude involvement of yeast proteins in this binding, the in vitro interaction of GST p55-IC fusion proteins, produced by bacteria, with the protein encoded by the 3 kB 55.11 cDNA {55.11-full), produced by transfceted HeLa cells, was examined. In this study the cDNAs for the full-length 55.11, FLAG-SS.II
(residues 309-900 of 55:11 encoded by the originally cloned partial cDNA and fused at the N
terminus with the FLAG octapcptide), and lucifcrase (control) were expressed in transfected HeLa cells and metabolically labeled W th (-ASS] Met and [355] Cys. The following proteins were fused with GST : full-length pS5-IC (GST-pSS-IC) and pSS-IC C-terminally uuncated up to amino acid 345 (GST-p55-IC345) to remo~~e most of the 'death domain' (see Table ~). GST alone served as a control. Lysates of the transfected cells were immunoprecipitated with antibodies against the 55.1 I protein when the full-lengnh 55.11 protein was used for binding the GST-fusion proteins, or with antibodies against the FLAG octapeptide when the FLAG-55.11 fusion product was used for binding the GST-fusion proteins. The proteins were analyzed by SDS-polyacrylart~ide gel electrophoresis (SDS-PAGE; 10°.o acrylamide), followed b5 autoradiography.
In Figs. 3A and B are shown reproductions of the autoradiograms of the above SDS-PAGE gels, in which Fig. 3 A depicts the binding of the full-length 55.11 protein (55.11-full) to the various GST-fusion proteins; and in which Fig 3B depicts the binding of the Flag-55.11 fusion product to the various GST-fusion proteins. In Fib. 3 A there is shown in the extreme right hand lane a control immunoprecipitate of lysates of cells transfected with only the full-length 55.11 and immunoprecipitated with the anti-55.11 antibodies (x.55.11 Abs). In Fig. 3B there is shown in the extreme right band lane a control immunoprccipitatc of lysates of cells transfected with only the FLAG-55.11 and immunoprecipitated with the anti-FLAG antibodies {otFLAG
Abs).
Thus, it is apparent from Figs. 3A and B that the protein encoded by the full-length 55.11 cDNA can be expressed in HeLa cells and it binds to fusion proteins that contained the full p55-IC (GST-p55IC) or a truncated p55-1C that lacked most of the death domain (GST-p55IC345) (Fig. 3 A). The full-length 55.11 protein did not bind to GST alone (control). Similarly, the HeLa ceU-expressed protein encoded by the initially cloned partial cDNA of 55.1 I in fusion with the FLAG octapeptide (FLAG-55.11) bound in vitro to GST-p55IC and GST-p55IC345, but not to GST (Fig. 3B). The above results also therefore provide additional evidence (see (b) above) that the 55.11 binds to a region of the pSSIC upstream of the 'death domain', i.e. in the region of the p55-IC that is more proximal to the transmembrane domain.
Moreover, the above study also demonstrates that, in accordance with the present invention, antibodies to 55.11 have been successfully produced {Fig. 3A).

4~
Le) Comparison of the deduced amino acid sequence of human SS.II to that of related ro~_teins present i tower organisms, and sequence features of the 55.11 rp otein As mentioned above, in accordance with the present invention, the full-length ;5.11 eDNA has been cloned and sequenced (see nucleotide sequence in Fig. 1 (a)) and the full amino acid sequence of 55.11 has been deduced from the cDNA sequence (see amino acid sequence in Fig. I(d)). Data bank (GenBanhThSBMBL DataBank) searches revealed that parts of the sequence of the human 55.11 cD\A (accession numbers T03659, 219559, and F091~8) and its mouse homologue (accession numbers X80422 and Z3I147) have alread~~ been determined during arbitrary sequencing of cDNA libraries. A cDNA sequence (accession number U18247) that encodes for a human protein of ~ 96 amino acids present in cultures of human hepatoma HC10 cells is similar to that of 55.11. This hepatoma protein, however, lacks an N terminal portion (amino acids 1-297) corresponding to that of 55.11 and also differs from 55.11 at the regions that correspond to residues 29%-3?7 and residues 648-G68 in 55.11. The searches of the data bank also revealed that proteins with very high sequence homolo~ry~ to 55.11 exist in Saccharomyces cerevisiae (yeasts), arahi~iup~is rhaliana (plants) and C'.aenorhahtlitis eleyanv (worms). Thus, 55.11 appears to fulfill an evolutionat5~ conserved function.
In the yeasts, there are two known proteins (the open readint frame ~C'HItG27c and SEN3) whose DNA
sequences resemble that of 55.11. The sizes of both are close to that of 55.11. YHR02?c is known only by the sequencing of a genomic clone while SEN3 has been cloned as a cDNA. The sites within 55.11 that are similar to those in SEN3 correlate to the sites of its similarity to YHR027c, although much more similarity is evident between X5.1 ! and YHIZ027c than between SS.I I and SEN3. The DNA sequence information available for the Aicthiclopsis thaliana and Caenorhahditis elegam proteins, although only partial, clearly shows that these proteins are as similar to 55.11 as the 'Y>$027c protein of yeast. The only one of these four proteins whose nature has been eluc'sdated so far is the yeast SEN3, whose homology to ~S.II
is limited. SElvT3 has been identified as the yeast ~qui~alent of the p112 subunit of an activator of the 20S
proteasome (the proteolytic core of the 2GS proteasome [Rechsteiner et al., 1993; Del\Sartino et al., 1994j) (M.R. Culbertson and M. Hockstrasser, personal communication).
1n Fig. 4 there is shown schematically a comparison of the deduced amino acid sequence of human 55.11 to that of the above-mentioned, related proteins present in lower organisms. In Fig. 4 the sequences that are compared are the sequences of amino acids predicted for : the 55.11 eDNA (sec Fig. 1 (d)); an open reading frame {YT-~t027c) within a cosmid derived from the 8th chromosome of Saccharnmvce.~ cerevisiae (nucleotides 21253-24234, accession number U10399); SEN3, the cDNA of a Saceharonryces cereui~iac protein (accession number L06321); a partial cDNA of a protein of the plant Arcrbidopsis thaliana (accession number T21500); and a partial eDNA of a proton of the nematode C.'aenorhahditi.s elegan.c (accession number D2?396). The 'KEKE' sequence in 55.11 is marked with a solid line and the sequence AI'AGS(x)gLL with broken Iines. The sequences were aligned using the 1?lLIrUP
and PRET?I'BOX programs of the CrCG package. Gaps introduced to maximize alignments are denoted by dashes.

rls regards the various sequence features or motifs present in the human 55.11 sequence the following has been obsewed : Consewed amino acid sequence motifs were not discerned within the protein encoded for by 55.11, except for a repetitive 'KEhE' sequence that extends between Lys 61.t and Glu 632 (underlined in Fig. 4). Such 'K,)rICE' sequences, which are present in many proteins, including proteasonal subunits and chaperonins, may promote association of protein complexes (Reaiini ct al., 1994). A sequence AYAGS(x)gLL appears twice in the 55.11 protein (at sites 479 590. see Fib. 4); no functional significance for this sequence has yet been described.
(f3 ~~guence features of the p5SI~ region invoived in binding to the ~~
~l~rot~in As described above (see (b) and (d)), the 55.11 protein binds to a region of the p55-IC between residues 243 and 308 (the N terminus of this binding site being between residues 243 and 266), this region being upstream ef the 'death domain' and more proximal to the transmembrane domain of the p55-T?v'F'-R. This region within p55-IC to which 55.1 l binds has a high content of proline, serine, and threonine residues. However, this region does not contain the I S RPM/ and RPIvL? proline-rich motifs present in several other cytokine receptors (OTleal and Y u-Lee, 1993). In the region that extends ben~~een residues 243 and 266, whose deletion abolishes the binding of p55-R to 55.11 (see (b) and (d) above and Table 3), two of the s~rinas and two of the threonines are followed by prolinc residues, which makes them potential sites for phosphorylation by MAP lcinase, Cl~C2, and other profine-dependent kinases (Seger and Krebs, 1995). Phosphorylation of this site in the receptors might affect its binding to the 55.11 protein.
In view of all of the aforementioned with regards to protein 55.11 and its binding to p55-IC it can be concluded that in accordance with the present invention, a new protein has been found which binds to a distinct region upstream to the 'death domain' of p55-IC. Such binding could affect TNF-mediated activities other than induction of cell death. The region to which 51.11 binds has previously been shown to be involved in induction of nitric oxide sy~nthase (Tartaglia et al., 1993), and appears to be involved in the activation of the neutral sphingomyelinase by ?NF
(Wiegmann et al., 1994). It is thus possible that association (binding) of 55.11 w7th the intracellular domain of p55-T:~IF-R (p55IC) affects or is involved in : (i) the signalinb for these above noted or other TNF effects, (ii) the folding or processing of the protein (as suggested by the similarity of 55.I 1 to a subunit of the 26S proteasome), or (iii) the regulation of the activity or expression of p55-TNF-R

if c 'on abili ~ of th i t ace lular domain f th 5'S T'V r . r ~5 n its capability to reuse cell death and other features and activities thereof. and a related ~as%APOl receptor's intra~~l~,lar docnain As set forth in Example 1 above, it was discovered that the intracellular domain of p55 TNF-R (p55IC) is capable of binding to itself, and further that fragments of p55IC, namely proteins 55.1 and 55.3, arc also capable of binding to p55IC.

' - CA 02490080 1995-05-11 ~7 It is know chat the binding of T1t to F5S T:~I'1:-R leads to a c~-tocidal effect on the cells carn~ing this receptor. Further, antaodies against the ewracellular domain or this receptor can themselves nigger this effect, in correlation with the effectivty of receptor cross-linking by them.
In addition, mutational studies (Tartaglia et al., (1.993); Brakcbusch et al., (1992)) showed that the function of the p5~-R depends on the integrity of its intracellular domain. It was therefore suggested that the initiation of signaling for the cytocidal effect of TNF
occurs as a consequence of association of two or more intraceltuiar domains of the p55-R (p55-IC), imposed by receptor aggregation. The results in accordance with the present invention provide further evidence for this notion, showing that e~;pression of the intracellular domain of the p55-R
within cells, without the transmembrane or intracellular domain, triggers their death. Such free intracellular domains of the p55-R are shown to self associate, which probably accounts for their ability to function independently of TNF. The fact that the si~~naling b; the full length p~5-R
does depend on TNF
stimulation is suggested to reflect acti~~yti(es~ of the transmembrane or e~ctracellular domain of the receptor which decrease or prevent this self association.
The ability of the intracellular domain of the p55-R (p55-IC} to self assaciate was found serendipitously, in the attempts to clone c~ector proteins which interact with this receptor (see Example 1 above). We applied for that purpose the above mentioned "two hybrid"
technique. In addition to the novel protein, 55.11 found to associate (bind) to the p55IC, it was also found that three other cloned HeLa cell cDN.~s contained cDNA sequences ancoding for parts of the inuaceIIular domain of the p55-R, implying that the p55-IC is capable of self association. Two of these clones were identical, containing an insert which encodes for amino acids 328-42f (designated as clone 55.1 encoding protein fragment 55.1 of the p55IC}. The third contained a Longer insert, encoding for amino acids 277-426 (designated as clone 55.3 encoding protein fragment 55.3 of the p55IC).
In addition, we assessed the 'tn_ vitro interaction between two bacterially produced chimeras of the p55IC, one, in which it was fused to the maltose binding protein {MBP) and the other in which is was fused to the ~,~lutathione-S-transferasc (GST). These chimeras were constructed, cloned and expressed by standard methods. Following their expression, the assessment of the self interaction of the p55-R intracellular domain (p55IC) by determining the interaction of the above bacterially-produced chimeric proteins GST-IC55 (Mr -SIkD) and MBP-IC55 (Mr - 6? l;D) with cacti other. Equal amounts of the GST-IC55 chimera (samples of lanes 1-4 in Fig. 5) or GST alone (samples of lanes 5-8 in Fig. 3) were bound to glutathione-agarose beads (Sigma) and were then incubated with the same amount of MBP-rC55 fusion protein in one of the following buffer solutions : = -(i) buffer I (20m1vI Tris-HCI, pH 7.5, IOUmM KCI, 2tnM CaCIZ, 2tnM MgCIZ, SmM
DTT, 0.2°.'o Triton X100*O.SmM PMSF, 5°fo Glycerol). This was done for the samples of Lanes 1 and 5 of Fig. 5.
(ii) buffer I containing SmM EDTA instead of M,gCIZ. This was done for the samples of Lanes 2 and 6 of Fig. 5.
* Trade-mark .~8 (iii) buffer I containing ~SOrrLli instead of 100mM KCI. This was done fo: the samples of Lanes 3 and 7 of Fig. 5.
(iv) buffer I containing 400m.1I inaead of 100mIVI KC1. This was done for the samples of Lanes 4 and 8 of Fig. 5.
After incubation with rotation for ~h at 4°C, the beads were washed with the same buffers and then boiled in SDS-PAGE buffer followed by electrophoresis by PAGE. The proteins on the gel were then Western blotted to a nitrocellulose membrane which was then stained with polyclonal antiserum asainst MBP. A reproduction of this stained Western blot is shown in Fig. 5, the samples in lanes 1-8 being those noted above.
From Fig. 5 it is apparent that the p55IC-MBP chimera bind to the pSSIC-GST
chimera (lanes 1-4) independently of divalent cottons and even at a rather high salt concentration (0.4M
KCI). Thus, it is concluded that the p~=IC is able to avidly self associate.
To evaluate the functional implications of the propensity of the p55-IG to self associate, we attempted to express the p55-IG within the cytoplasm of cells which are sensitive to the cytocidal effect of Th'F. Considering the possibility that the p55-IC wzlI
turn to be cvtotaxic, we chose to express it in an inducible manner, using the recently developed, tightly regulated tetracycline-controlled mammalian expression system (Gosscn and Boujard, 1992}. Expression of the p55-IC resulted in massi~~e cell death (Fig. 6, right panel). The dying cells displayed cell surface blabbing as observed in the killing of the cells by TNF. Transfection of the p53-IC
construct to the cells in the presence of tetracycline, which reportedly decreases the expression of pI~lO-3 regulated constructs by as much as IO$ fold, still resulted in some cell death, although significantly Iess than that observed in the absence of tetracycline (Fig 6, left panel). In contrast, cells transfected with a control construct, containing the lucipherase cDNA, showed no sums of death (results not shown).
?5 The ability of the p55-IC to trigger cell death, when expressed without the tra,nsmembrane or extracellular domains of the receptor, pro~rides further evidence for the involvement of this domain in signaling. Furthermore, it indicates that no other part of the receptor plays a direct role in such sirs~naling.. Studies of the effects of mutations, including those mutations studied in the present invention, on the function of the p55-IC, indicated that the region extendinb between amino acid residues 326 and 40? is most critical for its function. This rc~ton shows marked resemblance to sequences within the intracellular domains of two other receptors, evolutionarily related to the p55 TIFF-It- namely, the Fas receptor (Itoh et al., 1991; pohm et al., 1992), which can also signal for cell death and CD40 -a receptor (Stamenkovic et al., 1989) which enhances cell growth; this sequence therefore stems to constitute a conserved motif which plays some kind .5 of veneral role in signaling. Since it dots not resemble known motives characteristic of enzymatic activities, it seems plausible that it signals in indirect manner, i.e, possibly by serving as a dockinb site for signaling enzymes or for proteins which transmit stimulatory signals to them. The p55-IC, the Fas receptor and CD 40 can all be stimulated by antibodies against their cxtracellular domain.
Their stimulation could be shown to correlate with the ability of the antibodies to cross-fink the receptors. It therefore stems that the signaling is initiated as a consequence of interaction of two ~9 or more intracellular domains imposed bs- aggregation of the extracellular domains. Involvement of such interaction in the initiation osic.~naling of these receptors was also indicated by studies (Brahebusch et al., 1992) showing that cspression of receptors made nonfunctional by mutation of their intracellular domain, had a ''dominant negative" effect on the function of co-expressed normal receptors. Aggregation of the p= ~-R in response to TNF was suggested to occur in a passive manner, merely due to the fact that each of the TNF' molecules, which occur as homotrimers, can bind two or three receptor molecules. However, the findings of the present invention suggest that this process occurs somewhat differently.
The propensity of the p55-IC to self associate indicates that this domain plays an active role in its induced abgregation. More~~~er. this activity of the p55-IC seems to suffice for initiating its signaling, since when expressed independently of the rest of the receptor molecule, it can trigger cell death in the absence of T'.v'F or any other exterior stimuli.
Nevertheless, when expressed as the full length receptor, the p55-TNF-R does not signal, unless stimulated by TIvTF.
One must, therefore, assume that when activating the p55-TNF-R, TNp actuall3~
overcomes some inhibitory mechanisms, which prevent spontaneous association of the intracellular domains, and this inhibition is due to the linkage of the p55-IC to the rest of the receptor molecule. The inhibition may be due to the orientation imposed on the intracellular domain by the transmembrane and extracellular domain, to association of some other proteins with the receptor or perhaps just due to restriction of the amounts of receptors that arc allowed to be placed in the plasma membrane. Of note, this control mechanism should be rather effective, since according to some estimations, the binding of wen just one TN'F molecule to a cell suffices for the tribgering of its death.
Spontaneous signaling, independent of iigand can result in extensive derangement of the process controlled by this receptor. The best lmow~n example is the deregulation of growth factor receptors. h4utations due to which they start signaling spontaneously, for example those that cause them to aggregate spontaneously, play an important role in the deregulated growth of tumor cells. TNF effects, when induced in ehcess, are well known to contribute to the patholobry of many diseases. The ability of free intracellular domains (p55ICs) of the p$5-TNF-R to signal independently of TrrF may contribute to such excessive function. It seems possible, for example, that some of the cytopathic effects of viruses and other pathogens result, not from their direct cytoeidal function, buE from protcolytic detachment of the intracellular domain of the p55-TIVF-R
and the resulting TNF-like cytotoxic effect.
To further elucidate the re5rion(s) within p55IC which is responsible for its self association capability and hence its ligand-independent cell cytotoxicin~, and also to determine whether other related members of the TNF/NGF receptor family (e.g. FAS-R) also have invacellular domains with self association capabilities and ligand-independent effects, the following detailed study. was performed : ' . . ;

SG
~1 General Procedures and A4aterials (i) Two hybrid screen and two-hybrid d-~alactosidase expression test cDNA inserts, encoainL the p~5-IC and its deletion mutants, the Fas-IC and various other proteins (see Table 3), were cloned by PCR, either from the full-length eDNAs cloned previously in our laboratory, or from purchased cDNA libraries. ~i-galactosidase expression in yeasts (SFY526 reporter strain (Barrel et al., 1993)) transformed with these cDNAs in the pGBT-9 and pGAD-GH vectors (DNA binding domain (DBD) and activation domain (AD) constructs, respectively) was assessed by a liquid test (Guarente, 1983); it was also assessed by a fiber assay, yielding qualitatively the same results (not shown). Two-hybrid screening (Fields and Song, I9s9) of a purchased Gal4 AD-tagged HeLa cell cDNA library (Clontech, Palo Alto, Ca., U.S.A..) for proteins that bind to the intracellular domain of the pS5-R. (p55-IC), was performed using the HF7c yeast reporter strain according to the recommendation of the p:oducer. Positivity of the isolated clones was assessed by (a) prototrophy of the transformed yeasts for histidine when gown in the presence of 5 rn_'~~ 3-aminotriazole, (b) ~-galactosidase expression (c) specificity tests (interaction with SI\'F4 and lamin fused to Gal4 DBD).
(ii) In vitro self association of bacterialfv oroduced~55-IC fusionnrateins Glutathione S-transierase (GST) and glutathione S-transferase-p55-IC fusion protein (GST-p55-IC) were produced as described elsewhere (Frangiani and Neel.
19Q3;
Ausubel et al., 1994). Maltose binding protein (lt,~LBP) fusion proteins were obtained using the pMalcRI vector (New England Biolabs) and purified on an amylose resin column.
The interaction of the MBPP and GS? fusion proteins was investigated by incubating glutathione-agarose beads sequentially with the GST and N!$PP fusion proteins (5 pg protein I 20 ~tl beads; first incubation for 15 min, and the second for 2h, both at 4°C). Incubation with MBP
fusion proteins was carried out in a buffer solution containing 20 mM Tris-HCI, pH 7.5, 100 mM KCI, 2 mM
CaCl2, 2 mM
MgCl2, 5 mM dithiotreitol, 0.2% Triton X100, 0.5 rnM phenyl-methyl-sulphonyl-fluoride and 5%
(v/v) glycerol or, when indicated, in that same buffer containing 0.4 M KCI, or 5 mM F-DTA
instead of MgCl2. .Association of the MBP fusion proteins was assessed by SDS
polyacrylamide gel electrophoresis (10% acrylamide) of the proteins associated with the glutathione-agarose beads, followed by Western blotting. The blots were probed with rabbit antiserum against MBP
(produced in our laboratory) and with horseradish-peroxidase-linked goat-anti-rabbit immunoglobulin.
(iii) induced expression in kleLa cells of the",~35-R and fragments thereof HeLa cells expressing the. tetracyciinc-controlled transactri~ator developed by Gossen and Bujard (the HtTA-1 clone (Gossen and $ujard, 199?)), were grown in Dulbecco's modified Eagle's mediurtL containing 10% fetal calf serum, 100 u/mI
penicillin, _ 104 ~tc~/ml streptomycin and 0.5 mg/ml neomycin. cDN A inserts encoding the p55-R or pans thereof were introduced into a tetracycline-controlled expression vector (pL~HD 10-3, kindly provided by H.
Bujard). The cells wore transfected with the expression construct (5 ug DNAIG
cm plate) by the calcium phosphate precipitation method (Ausubel et al., 1994). Effects of transient expression of the transfected proteins were assessed at the indicated times after transfcction in the presence or WO 95131544 pCTJUS95105854 sl absence of tetracycline (1 uglml). Clones of cells stably transfected with the human p55-IC cDlv'A.
in the pUl-iD 1 Q-3 vector were established by transfecting the cDNA to HtTA-1 cells in the presence of tetracycline together W th a plasmid conferring resistance to hygromycin, followed by selcaed for clones resistant hygromycin (200 p~rJml). Expression of the eDNA
was obtained by ~ removal of tetracycline which was otherwise maintained constantly in the cell growth medium.
(iv) Assessment of TNF-like effects tr~~~ered b~nduccd expression of the o5~-R
~d fra~,ments thereof Effects of induced expression of the receptor and of TNF on cell viability ~werc assessed by the neutral-red uptake method (Wallach, 1984). Induction of IL-8 gene expression was assessed by Northern analysis. RNA was isolated using TRl REAGENT
(Molecular Research Center, Inc.), denatured in formaldehyde/formamide buffer, electrophorcsed through an agaroselformaldehyde gel and blotted to a GeneScreen Plus membrane (Du Pont) in IOxSSPE
buffer, using standard techniques. Filters were hybridized with an IL-8 cDNA
probe (Matsushima et al., 1988), nucleotides 1-392). radiolabeled by the random-prime L-it (Boehringer Mannheim Biochemica, Mannheim, Germany) and washed stringently according to the protocol of manufacturer. Autoradiography wzs performed for 1-2 days.
(v) Assessment of TNF rece~to; ex~res~ion TNF receptor expression in samples of 1x106 cells was assessed by measuring the binding of TNF, labeled with IZ$I by the chloramine-T method, as previously described 30 (Holtmann and Vfallach, 1987). It was also assessed by ELISA,, performed as described for the quantification of the soluble TNF receptors (Aderka et al., 1991), except for the use of RIPA
buffer (10 mM Tris-HCI, pH 7.5, 150 mM NaCI. 1% NP-40, 1% deoxycholate, 0.1°lo SDS and 1 mM EDTA) to lyse the cells (70 uUlOb cells) and to dilute the tested samples.
The soluble form of the p55-It, purified from urine, screed as the standard.
b M tational an 1 sis of t tra el ula d main o a 5 - ~- to d termine t a regions of the p55-IC involved in itc self accoci~,i~
As noted above, p55-IC can self associate and trigger cytotoxic effects on cells, and there are portions of the p55-IC, which themselves were capable of binding to the full-length p55-IC. In particular, one of the portions of the p55-IC (designated as protein fragment 55.1 in Example 1 above) was identified that was capable of binding strongly to the full length p55-IC, this portion was sequenced and was observed to contain the amino acid residues 328-426 of the p55-TNF-R
which are within the p55-IC. It has further been discovered (see below) that the above portion, protein fragment 5~.1, is itself capable of self association and of triggering cytoto~ac effects on cells. Hence this portion of the p55-IC has been called the 'death domain', and is located in the rer.~'on beEween amino acid residues 328-426 of the human p55-R, most likely consisting of amino acid residues between about residue 328 and 414 thereof.
The fact that the 'death domain' in the p55-IC self associates was found by happenstance.
On screening a HeLa cell cDNA library by the two-hybrid technique (see Example 1 above) for proteins that bind to the intracellular domain of this receptor, we detected among the cDNAs whose products bound specifically to the intracellular domain-G.~~L,4 DBD
fusion-protein, several clones (e.g. 55.1 and 55.3) that themselves encoded for parts of the p55-R
intracellular domain (p55-IC; marked with asterisks in Table 3).
Applying the two-hybrid test to evaluate the ea-tent of specificity in the self association of p55-IC and to define more accurately the region involved Ied to the following findings (Table 3) : .
S (a) The self association of p55-IC is coned to a region within the 'death domain'. Its N terminus is located between residues 328 and 344 and its C terminus, close to residue 404, somewhat upstream of the reported G terminus of this domain (residue 414). (b) Deletion of the membrane-proximal part of p55-IC upstream of the 'death domain' enhanced self association, suggesting that this region has an inhibitor effect on the association. (c) Mouse p55-1C self associates, and also i0 associates with the 'death domain' of human p55-R. (d) examination of the self association of the intracellular domains of three other receptors of the TNF/NGF receptor family : Fas/APO1 (FAS-R), CD40 (Fields and Song, 1989) and the p75 T~'F receptor (Smith et al..
1990), showed that Fas-IC, which signals for cell death by a sequence motif related to the p55-R
'death domain', self associates, and associates to some extent with the p55-IC. However, CD40-IC, that provides IS gowth stimulatory signals (even though also containins a sequence resembling the 'death domain'), and p75-IC, that bears no structural resemblance to p55-IC, do not self associate, nor do they bind p55-IC or Fas-IG.

TABLE 3. Self association of the intracellular domains of p55-R and Fas/AP01 within transformed yeasts : assessment by a two-hybrid ~i-galactosidase expression test.
U

C I ( I o 0 0 o a o I

C~

I I I I I I I I N I i cn ..
.U

O =_~ I I I I l ! ! I I i =~

... _ ~ _ ... .._ ' ' = 1 I I I I I I i ( I I

_ _ O y I I I I I I ~ o o I o 0 .

c:.

~U

.

I I I i I N ~ a I I I

O

N

N
f'f I I

L. e-W

< N ~ I 1 1 I I I I I I I I

t,, h i I I I 1 I I I I I I

.

' ~' I I I I i I ( ( I 1 I
xt c ...;..~~ r~
L

O G ~ C O O

O

P7 , . , . .
Q

00 V M O O C1 ~.
V

n n; p N

N
'et .

tV O N O O V ...0 I O .' O V1 1~7 N

N v Q

U

U U U
' '~ - ..
~ U
Z
'~

O In H ~

. G L G

.
V V V IA ': ~ Q G1 V _ < N N N 'C N ::~ V tJ ~ C
"~ "", ' .... G""

r ef Q ~ M M
. t = H H U P.
r_ t 1 1 1 " ..
~ t r o N ~ o o ~"'~~ ~ z ~ v o ~ ~ - N - W ~ .
~ N .i r N f Ih .. .. .-.~. ..
. v S

9Zb ;

=

a , s~ o c~ . . .
, tc~ D ;
g a _, e , WO 95131544 pCTNS95105854 >4 Table s abo~~e shoms the quantitati~~e assessment of the interaction of Gal4 hybrid constructs encompassing the follow n~- proteins : the intracellular domain of human p5>-R and its various deletion mutants (residues numbered as in (Loetscher et al., 1990));
the intracellular domains of mouse p55-R (residues 334-454, numbered as in (Goodv~~in et al., 1991)); mouse S FasIAP01 f,Fas-IC, 166-306, numbered as in (Watanabe-Fukanaga et al., 1992)); human CD40 (CD40-IC, 216-277, numbered as in (Stamenkovic et al., 1989)): and human p7S
TI~TF receptor (p?5-IC, 287-461, numbered as in (Smith et aL, 1990)). SNF1 and S'NF4 were used as positive controls for association (Fields and Song, 1989), and lamin as a negative control (Barrel et al., 1993). Proteins encoded by the Gal4 DBD constructs (pGPT9) are listed vertically; those encoded by thr Gal4 AD constructs (pG.AD-GH), horizontally. The two deletion mutants denoted by asterisks were cloned in a two-hybrid screen of a HeLa cell cDNA library (Clontech, Pato Alto, Ca., U.S.A.) using p55-IC cloned in pGBT9 as "bait". In that screen, four of about 4Yt 06 cDNA clones examined were positive. Three of these clones were found to correspond to parts of human ps5-R cDNA (two were identical, encoding residues 328-426 and one encoding residues 277-426). The fourth was found to encode an unknown protein. The (3-galactosidase expression data are averages of assays of tv~~o independent transformants and are presented as amount of (3-galactosidase product; (a unit of activity being defined as OD42p times 103 divided by OD6pp of the yeast culture and reaction time, in minutes). The detection limit of the assay was 0.05 units.
Variation between duplicate samples were in all cases Icss that 25% of the average (not tested}.
An in vitro test of the interaction of a p55-IC-glutathionc-S-transferase (GST) bacterial fusion protein with a p55-IC-maltose binding protein (IvlBP) fusion protein confirmed that p55-R
self associates and ruled out involvement of yeast proteins in this association (see above}. The association was not affected by increased salt concentration, nor by EDTA (see above).
To evaluate the functional implications of the self association of the death domain, we examined the way in which induced expression of p55-R, or of parts of it, affect cells sensitive to TNF cytotoxicity. The results of this analysis art set forth in Fig. 7 which depicts the li~and independent tribgering of a cy~tocidal effect in HeLa cells transfected with p55-R, its intracellular domain (p55-IC) or parts thereof (including the 'death domain').
In Fig. 7 there is shown schematically, the various DNA molecules encoding the different types of TNF receptors included in the vectors with which the HeLa cells were transfected (extreme left hand side of Fig. 7); and the expression (left and middle bax graphs) and the viability (right bar graph) in HeLa cells expressinfi transiently the various full-length g55-R (p55-R), p55 IC or parts of p55-IC or, as a control, luciferase (LUC) (each being depicted at the extreme left side of Fig. 7), using a tetracycline-controlled expression vector. The open bar graphs (left, middle and right) represent cells transfected in the presence of tetracycline (1 pg~ml), which inhibits expression; and the filled bar graphs (left, middle and right) represent Celts transfected in the absence of tetracycline. Tl~t'F receptor expression was assessed 20h after transfection, both by EL1SA, using antibodies against the receptor's e~,-tracellular domain {see schematic illustration on the left side of Fig. 7), and by determining the binding of radiolabeled TNF
to the cells (middle).
The cytocidal effect of the transfected proteins was assessed 4Sh after transfection. Daia shown WO 95!31544 CA 02490080 1995-05-11 pCT~595/05854 $$
are from one of three experiments v~th qualitatively similar results, in which each construct was tested in duplicate. ND - not determined.
Thus, from Fig. 7 it is apparent that by using an expression vector that permits strictly controlled expression of transfected cDN As bS- a tetracycline regulated transactivator (Gossen and Bujard, 1992), a mere increase of p~5-R expression in HeLa cells by expression of transiently transfected cDNA for the full-lengnh receptor resulted in quite e.~ctensive cell death. An even greater cytotoxicity was observed when expressing just ps5-IC. Significant cytotoxicity was also observ,~ed when expressing just a part of p5$-~C comprising essentially the 'death domain' (residues 328-42b) in the HeLa calls. On the other hand, expression of parts of pSS-IC
that lacked the 'death domain' or contained just part of it (or expression of the luciferase gene, used as an irrelevant control) had no effect on cell viability. The cytotoxicity of p~3-IC was further confirmed using cells stably transformed with its cDNA; these cells continued to grow when p5$-1C expression was not induced, but died when p55-IC was e.~cpressed (see above).
(~,~ Other eff'~, of thg intraceliul r domain of the p55-TNF-R
To examine whether other activities of TNF are triggered by the self association of the intracellular domain, including the 'death domain' thereof, we examined tb;e affect of increased e~,-pression of the full-length receptor (p55-R) and of the expression of the intracellular domain of the receptor (p55-IC), on the transcription of irncrlculdn 8 (IL-8), known to be activated by TNF
(Matsushima et al., 1988). The results are show in Fig. 8, which depicts the ligand-independent induction of iL-8 gene expression in HeLa cells transfected with p55-R or p55-IC, using a teuacycline-controlled construct (see also 'General Procedures and Materials' and Example 1 above). In panel A of Fig. 8 there is shown a reproduction of a Northern blot representing the Northern blot analysis (see 'General Procedures and Materials' above) of RNA
{7 ~g/lane) extracted from HeLa (HTta-1) cells, untreated ('control) or treated ('TNF') with TNF (500 l.ilml for 4h), or the HTta-1 cells 24h after transfection (in the presence or absence of tetracycline) with p55-IC ('p55-IC'), the p55-R ('p55-R~, or luciferase ('Luc') cDNA. In panel B
of Fig. 8 there is shown a reproduction of a Northern blot representing the mcthylcne blue staining of 18S rRNA in each of the samples shown in panel A of Fig. 8.
Thus, as is apparent from Fig. 8, transfection of HeLa ctlls with a tetracycline-controlled construct encoding the p5~-R cDNA induced IL-8 transcription. An even stronger induction was observed in cells transfected with the cDNA for p55-IC. In both cases, the induction occurred only when tetracycline was excluded from the cell growth medium, indicating that it occurs as a consequence of expression of the transfected p5.5-R or p55-IC. Transfection with luciferase eDNA, as a control, had no effect on Zh-8 transcription.
3 5 Accordingly, from the above results (Fig. 8), it appears that a mere increase in p55-R
expression, or even expression of just the inuacellular domain (p55-IC) thereof is sufficient to trigger, in a ligand (TNF)-independent fashion, cytotoxicity and other effects as well, including that of an increase in the ea-pression of the IL-8 gene w7thin.cells. The triggering of these effects is most likely due to the self association of the intracellular domain of the p55-R (p55-IC). As is set an f~nh a~~ve. it annear~ that. lln~r celf as~nciation of the n5~-1C. the 'death domain' thereof is primarily responsible for signaling the induction of the intracellular processes leading to the triggering of cytotvxicity within the cells, u~hilst the other effects, e.g the sitnalinw leading to the induction of 1L-8 gene expression, are likely due to other regions of the p55-IC as well, following the self association thereof. It is therefore possible that different regions of the p55-IC are responsible for the different T1'F-induced effects (e.g. cytotoxicity, IL-8 induction) within cells, these effects being a consequence of the intracellular si~~naling upon self association of the p.SS-IC.
T'he fact that the pS5-IC, can induce in a ligand (TNF)-independent fashion, the triggering of other intracellular effects e.g. IL-8 induction, means that the p55-IC or specific portions thereof may be used as a hie;hly specific tool for bringing about such effrects in cells or tissues that it is desired to treat, without the need for treating such cells or tissues with T=VF. In many pathological conditions (e.g. malignancies), treatment with TNF, especially at high dosages can lead to undesirable side-effects due to the number of intracellular effects induced systemically by ThtF following its binding to its receptors. By way of the discovery in a:.cordance with the present invention that the p55-IC can mimic specific other TNF-induced effects (besides cytotoxicity), e.g. II,-8 induction, opens the wa}~ for introducing in a cell- or tissue-specific manner, p5s-IC or specific portions thereof, which will be capable of signaling for the induction of specific desired intracellular effects, e.g. IL-8 induction, and thereby overcome the systemic side-effects often observed during TNF treatment.
~~gand-indsnendent triggering, of cytocidal eff~ct~ in H~La cells bY the intracellular domains and. thc'death domains' thereof of p5S TIfF R and FAS-R r[Fas/A~O~,~
As regards the cytotoxic activity of the intracellular domains of the p55 TNF-R and FAS-R (p55IC and FAS-IC) it has now also been further elucidated that both the p551C, its 'death domain' (p55DA) and the FAS-IC are capable of a Iigand-independent triggering of a cy~tocidal effect in HeLa cells. In this study, HeLa cells were transfected with expression vectors containing various constructs of either the full-length p55-TIv'F-R, portions thereof including the p55IC and p55DD or the FAS-IC. In one set of experiments HeLa cells were co-transfected with constructs containing the p55 TNF-R (p55-R) and the FAS-IC (for details of the constructs, their preparation, etc. see above). The results of this study arc depicted in Fig. 9 ( A and B j, wherein in both Fig. 9A and B the constructs used for transfecting the HeLa cells are shown schematically in the left hand panels; the results of the TNF or FAS receptor expression are.
shown graphically in the two middle panels (second and third panels from the left); and the results of transfected cell viability are shown gaphicallv in the right hand panels. In Fig. 9A there is shown the results of transfected HeLa cells transiently expressinb the full-length p55-R, p55-IC or pans thereof, or as 3~ a control, lucifcrase (LUC), in all cases using a tetracycline-controlled expression vector. In Fig.
9B there is shown the results of transfected HeLa cells transiently expressing FAS-1C alone or together with the p55-R, using a tetracycline-controlled expression vector. In thegraphic representation of the results in Fig. 9A and B, the open bars represent cells transfected in the presence of tetracycline (1 itg/ml), which inhibits expression, and the closed bars represent cells transfected in the absence of tetracycline. TNF receptor expression was assessed 20h after Si transfection, both by ELISA usinL antibodies against the ea-tracellular domain of the receptor {see left hand ~anelsj, and by determining the binding of radiolabeled TNF to the cells (middle panels).
The cytocidal effect of the transfected proteins was assessed 48h after transfection. The data shown are from one of three experiments with qualitatively similar results in which each construct was tested in duplicate The designation ND' in Figs. 9A and B means not determined. From the results shown in Figs. 9A and B it is apparent that expression of only the p»IC results in even greater cynotoxicity. Sisazificant cwotoxicity also occurs when expressing just the death domain (pSSDD), In contrast, expression of parts of p55IC lacking the death domun or containing only part thereof, had no effect on cell viability. Expression of the FAS-IC did not result in significant cytotoxicity, yet it significantly enhanced the cyotoacit5~ of co-expressed pS5-R.
~ 'AMPL ~,3 Additional proteins capable of binding to the intracellular domains of p~5 ~'1~F-R or FAS-R
I S Using the same approach and technology set forth in Example 1 above;
'three more proteins have been isolated and identified which arc capable of binding to the pSSIC or FAS-IC.
In Fins. 10-12 there is shown schematically the partial and preliminary 'nucleotide sequence of cDNA clones, called F2, F9 and DD 11, respectively.
Clones F2 and F9 were isolated by screening a marine (mouse) embryonic library using the marine FAS-IC as "bait". In Fig. 10 there is shown schematically the partial nucleotide sequence from iha F2 cDNA that has been sequenced. 1n Fig. 11 there is shown schematically the partial nucleotide sequence of 1724 bases from the F9 cDNA that has been sequenced.
Analysis of the binding capability of the protein encoded by clones F2 and F9 (F2 and F9, respectively) has shown that 2S {a) F2 interacts strongly with human pSSIC and pSSDD and with marine FAS-IC, while it interacts weakly with non-rele~-ant {control) proteins SNF1 and Lamin as well as relevant protein, human FAS-IC.
{b) F9 interacts strongly with human p5S-IC and marine FAS-IC, while it interacts weakly with human FAS-IC (relevant protein) and irrelevant proteins SNF 1 and Lamin.
(c) Neither F2 nor F9 interacted at all with human p75IC, pGBT9 (empty bait v ector), or human CD-40.
Further, from 'Gene Bank' and Protein Bank' searches it was revealed that F2 and F9 represent new proteins.
Thus, F2 and F9 represent new proteins having binding specificity for both FAS-IC
and pSSIC.
Clone DD11 was isolated by screening a human HeLa library using the human p55D17 as "bait". In Fig. 12 there is shown schematically the partial nucleotide sequence of 42S bases from the DD11 cDNA that has been sequenced.
The DD11 clone has an approx. length of 800 nucleotides. The full length of the transcript .1 f1 ie 41~ ny~ 1 7 t~A rant ant 1,?t~i~~ ~P n nfnl~Pr~ yqine t~A In ..
~,,~ol..esc of the ri:..r .. , . ~:~.'. t. C. . . . : C. . . . ~ . C . r_ _ .. _ r1!i _. -WO 95!31544 PCT/US95/05854 capability of the protein encoded by clone DD 11 has shown that DD11 interacts strongly with the pSSDD (a.a. 326-414) (see Fig. 9) and does not interact with deletion mutants of this domain, e.b.
a.a. 326-40.t. DD 11 also interacts with mouse and human FrIS-IC and to some extent also with ):.amin. DD 11 does not interact at all with SN'F1 nor with pGBT9 {empty bait vector). DD 11 is also not found in the 'Gene Bank' and Protein Bank' databases. Thus DD11 represents a pS5 l;C
(pSSDD) and FAS-IC specific binding protein.
x 1 4 construction of soluble dimeric TNF receptors ~ 0 Based on the findings set forth in Example 2 above, that the intracellular domain of the p55-R (p55-IC) and a portion thereof (che'death domain'); and that the intracellular domain of the FaslAP01 and a portion thereof (also called the 'death domain') which resembles the p55-IC
'death domain', are capable of self association, it is possible to construct new TNF receptors which are capable of self association (aggregation) and which are soluble.
Such TNF receptors 1 S will be fusion proteins having essentially all of the ex-cracellular domain of the pSS-R fused to essentially all of the intracellular domains or 'death domains' thereof of the p~~-R or Fas/APO1.
Thus, such fusion cc:nstructs n~iil be devoid of the transmembranal domain of the p55-R (or FASIAPO1) and hence wilt be soluble. Moreover, by virtue of the self association capability of the intracellular domains or 'death domains' thereof these fusion constructs will be capable of 20 oligomerization to provide at least dimers (and possibly also higher order multimers) of the p55-R Consequently, such dimeric TIv'F receptors (p55-R) will be capable of binding to at least two TNF monomers of the naturally-occurring Tr'F homotrirner to provide a soluble TNF receptor which binds more avidly to its ligand (homotrimeric TNF).
Accordingly, at least four types of p55 TNF receptor fusion proteins will be constructed 25 each of which will be capable of oligomcrization and will be soluble (i) A fusion product between the extracellular domain of pSS-R (EC55) and the intracellular domain of p55-R (p~5-1C);
(ii) A fusion product between the EC55 and the 'death domain' of p55-IC
(DD55);
(iii) A fusion product between the ECSS and the intracellular domain of Fas/AP01 30 (ICFAS); and (iv) A fusion product between the EC55 and the 'death domain' of ICFAS
(DDFAS).
In each of the above fusion proteins the T1VF monomer binding capability is provided by the EC» portion while the oligomerization (or at least dimerization) of each kind of fusion 3~ protein is provided by its'taif rer~~on being any of the p55IC, DDSS, ICFAS
or DDFAS portions.
For construction of the above fusion proteins, standard techniques of recombinant DN.A
technology will be employed that are now well established in the art {see for example Sambrook et al., (1989) Molecular Cloning : A Laboratory h4anual, Cold Sprint Harbor Laboratory Press, Cold Spring Harbor, N.Y.). Briefly, any suitable bacterial, bacteriophage, or animal virus 40 expression vector (cloning vehicle or plasmid designed for expression of the inserted DNA of VVO 95!31544 CA 02490080 1995-05-11 PCTlUS95105854 $o choice) may' be employed into which will be inserted in one or more stages the DNA encoding the ECSS and one of the 'tails' being the p55-1C, DD55, ICFAS or DDFAS. The so-inserted DNA
encoding each of the fusion proteins will be placed under the control of the carious expression control sequences of the clotting vehicle or plasmid such as promoters, ribozyme binding sites, transcriptions) factor binding sites, etc. These expression control sequences will be chosen depending on the type of expression vector chosen and hence the type of host cell {eukaryotic or prokaryoric) in which it is desired to express the fusion proteins of the invention. Preferred host cells (and hence expression vectors) are eukaryotic, in particular, mammalian.
The DNA molecule encodins each of the above noted fusion proteins will be prepared and inserted into the expression vector by the followinb procedure (a) Firstly, a set of oligonucleotides for use in PCR will be constructed by standard means, the oligonucIeotides being 1) ACC ATG GGC CTC TCC ACC GTG (ECSS, sense) 2) ACGC GTC G AC TGT GGT GCC TGA GTC CTC (ECS S, antisense) 3) ACGC GTC G:~C CGC TAC C.4A CGG TGG AAG (ICSS, sense) 4) TCA TCT GAG A.4G.ACT GGG (ICSS, antisense 5) ACGC GTC G.~C AAG AGA AAG GAA GTA CAG (IC FAS, sense) 6) CTA GAC C 4~. GCT TTG GAT (IC FAS, antisense) 7) ACGC GTC GAC CCC GCG ACG CTG TAC GCC tDD3S, sense) 8) ACGC GTC GAC GAT GTT GAC TTG AGT AAA (DD FAS, sense) (b) Plasmids containing the cloned full-length pS5-R and FasJAPOI receptors which we have in our laboratory (see also co-pending EPSG8925 and Examples 1-3 above) will be subjected to the following manipulations to yield the DIVA fragments encoding each of the fusion proteins, which DNA fragments are then ligated into the above noted expression vector of choice (i) To produce the ANA fragment coding for EC55 which is a component of all 4 fusion proteins, PCR is performed on a plasmid bearing cDNA of human p55 using the about oligonucleotide nos. 1 and 2 (size of fragmtent 640 bp).
(ii) To get a fusion product ECSS-ICSS, PCR is performed on a plasmid bearinb cDNA for human pSS using oligonucleotide nos. 3 and 4, to obtain a DNA
fragmont coding for IC55 (size 677 bp} which is then mixed with EC55 digested by Sal i and ligated by blunt end ligation into any expression vector for mammalian cells under the control of an appropriate promoter. The orientation of the inserted EC55--ICSS in the vector is verified by restriction digestion and by sequenang.
{iii) To get a fusion product EC55-IC FAS, IC FAS is produced by PCR an a 3 S plasmid with cDNA for FAS using olieonucleotide nos. S and 6, to obtain a fracmtent (size 448 bp) which is then cut by Sal I and mixed with EC55 cut by SaII, and subseducntly is blunt ligated into a manunalian expression vector under the control of an appropriate promoter. The orientation of the inserted ECSS--IC FAS in the vector is verified by restriction digestion and by sequencing. ;

(iv) To get a fusion product EC55--DDS, a DNA fragment is produced 7th the DD55 sequence by PGR in cDNA for human p5~ using oligonucleotide nos. 7 and 4.
The product with a size of 314 by is cut by SaII and mixed with ECSS cut by SaII, and subsequentl~~ blunt ligated into the mammalian expression vector. Orientation of the inserted EC55--DD55 in the 5 vector is vecificd by restriction digestion and by sequencing.
(v) To get a fusion product ECSS--DD FAS, a DNA fragnent with DD FAS is produced by PCR on cDNA for FAS usinz oligonucleotide nos. 6 and 8. The product with a size of 332 by is cut with SaII, and mixed with EC55 cut by Sal I and subsequently blunt ligated into the mammalian expression vector. Orientation of the EC55--DD FAS is then verified by 10 restriction digestion and sequencing Once the above expression vectors have been constructed, they will then be introduced by standard methods into suitable mammalian cells (e.g. Chinese Hamster Ovar;~
(CHO) or TTonkey Kidney (COS) cells) for the purposes of expression. The so-expressed fusion proteins will then be purified by standard methods (see co-pending EP308378; EP398327; and EP568925). The 15 purified fusion proteins v~711 then be analyzed for their ability to oligomerize (and the extent thereof, i.e. whether they form dimers or higher order multimers) and for their ability to bind TNF
(aad the affinity or avidiy of binding thereof).
za le 5 20 Constr ction of soluble dimeric FacIAP01 receptors In a similar fashion to that set forth in Example 4 above, it is possible to produce the following four kinds of Fas/AP01 fusion products, each of which will be capable of oligomerization and will be soluble (i) Fusion product betvvecn the extracellular domain of Fas/AP01 (EC FAS) and 25 the intracellular domain of p55-rC;
{ii) Fusion product between the fiC FAS and the 'death domain' of p55-IC
(DD55);
(iii) Fusion product bctv~~een the EC FAS and the intracellular domain of FasIAP01 (IC FAS); and 30 (iv) Fusion product between the EC FAS and the 'death domain' of IC FAS (DD
FAS).
In each of the above fusion proteins the FAS ligand binding capability is provided by the EC FAS portion, while the oligomerization (or at least dimerization) of each kind of fusion protein is provided by its 'tail' repon being any of the p5~-IC. DD55, IC FAS
or DD FAS
35 portions.
The construction of the DNA fragments encoding the above fusion proteins and expression vectors containing them v~~ill be as detailed in Example 4, except different appropriate olibonucleotides {not shown) will be used for the preparation of the EC FAS
fragment to be ligatcd to any of the above noted 'tail' regions. Subsequently, the expression vectors will be introduced into the suitable host cells. and the resulting expressed fusion proteins W II be purified WO 95/31544 pCT/US95/05854 and tested for their ability to oligomerize (and the extent thereof, i.e, whether they form dimers or higher order multimers) and for their ability to bind the FAS ligand (and the affinit}~ or avidity of binding thereofj.
S Example G
Construction of soluble oligomeric'mixed' TNP'fE~S._xeoento~s To prepare oligomcric receptors having 'mixed' aflanity, i.e. affinity for both ThIF and the FAS-R ligand, the above-mentioned (Examples 4 and 5) fusion products may be utilized in the followins procedure i) Providing a fusion product as set forth in Example 4, which contains the e~,-tracellular domain of a TNF-R (p75 TNF-R or p55 TiVF-R) fused to any one of : the p55 IC, FAS-1C, p55 DD or FAS DD;
ii) Providing a fusion product as set forth in Example S, which contains the extracellular domain of Fas-R fused to any one of : p53 IC, FAS-IC, p55 DD or FAS-DD; and iii) mixing any one of the fusion products of i) with anv one of the fusion products of ii) to provide a new dimeric (or higher order oligotneric) receptor which has both the extracellular domains of a Tr'F-R and FAS-R that are joined by their -IC or -DD regions.
In the above procedure the fusion products of i) and ii) may be provided separately, namely, from their purification from transformed cells in which they were produced, and then mixed in vitro to obtain the mixed affinity receptors. Alternatively, the host cells may be co transfected with vectors carrying sequences encoding both types of fusion products, in which case, the mixed affinity receptors may be obtained directly from the co-transfected cells. The actual oligomerization of the fusion products into oligomeric receptors may take place within the cells or during or following the purification procedure to obtain the fusion products expressed in the cells. To specifically select for the mixed affinity receptors any standard method may be utilized, for example, affinity chromatography procedures in which antibodies against the TNF-R
and FAS-R extraceilular domains are used in sequential chromatographic steps to scf~ct for thQSe receptors having; both types of extracellular domain.

REFERENCES
Aderka, D., Englemann, H., Hornik, V., Skornick, Y., Levo, Y., Wallach, D. and Kushtai, G.
( 1991 ) Cancer Res. 51 5602-5607.
Baens et al. (1993) Genomics x.6_:214-218.
Barinaga, M. (1993) Science X621512-4.
Bartel, P.L., Chien, C.T., Sternglanz, R. and Fields, S. (1993) Bio Techniques 14. 920-924.
Berger, J., Hauber, J., Hauber, R., Geiger, R. and Cullen, B.R. (1988) Gene 66 1-10.
Beutler, B. and Cerami, C. {1987) NEJM, 316:379-385.
Boldin, M.P. et al. (1995) J. Biol. Chem. X70 337-341.
Bollon, D.P. et al. (1980) J. Clin. Hematol. Oncol. 0 39-48.
Botstein, D. et al. (1982) Miami Wint. Smyp. 9 265-274.
Brakebusch, C. et al. (1992) EMBO J., 11:943-950.
Broach, J.R. (1981) in : The Molecular Bioloev of the Yeast Saccharomyces :
Life ~vcle and Inheritance. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 445-470.
1 ~ Broach, J.R. ( 1982) Cell 2s 203-204.
Brockhaus, M. et al. (1990) Proc. Natl. Acad. Sci. USA, 87:3127-3131.
Cantor, G.H. et al. (1993) Proc. Natl. Acad. Sci. USA~O_:10932-6.
Chater, K.F. et al. (1986) in : Sixth International Symposium on Actinom~etales Biology Akademiai Kaido, Budapest, Hungary, 45-54.
Chen, C.J. et al. (1992) Ann N.Y. Acad. Sci. 660:271-3.
Cheng, J., Zhou, T., Liu, C. Shapiro, J.P. Brauer, M., Kiefer, M.C., Barr, P.J. and Mountz, J.D.
(1994) Science 263. 1759-1762.
Crisell, P. et al., (1993) Nucleic Acids Res. (England) 21 (22):5251-5.
Crowe, P.D. et al., (1994) Science, 264:707-709.
Current protocols in molecular biology (Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., Albright, L.M., Coen, D.M. & Varki, A., eds.), (1994) pp. 8.1.1-8.1.6 and 16.7-16.7.8, Greene Publishing Associates, Inc. and Wiley &
Sons, Inc., New York. _ DeMartino, G.N., Moomaw, C.R., Zagnitko, O.P., Proske, R.J., Chu-Ping, M., Afendis, S.J., Swaffield, J.C. and Slaughter, C.A. (1994) J. Biol. Chem. 26Q, 20878-20884.
Dicks, W., Wirth, M. and Hauser, H. (1993) Gene 128_ 247-249.
Endo, H., Akahoshi, T., Nishimura, A., Tonegawa, M., Takagishi, K., Kashiwazaki, Matsushima, K. and Kondo, H. (1994) Clin. Exp. Immunol. 96 31-35.
Engelmann, H. et al. (1990) J. Biol. Chem., X65:1531-1536.
.35 Ferrick, M.R., Thurau, S.R., Oppenheim, M.H., Herbort, C.P., Ni, M., Zachariae, C.P., Matsushima, K. and Chan, C.C. (1991) Invest. Ophthalmol. Vis. Sci. 32 1534-1539.
Fields, S, and Song, O. ( 1989) Nature, 360:245-246.
Frangioni, J.V. and Neel, B.G. (1993) Anal. Biochem. 210. 179-187.
Glick, B.R. (1987) J. Ind. Microbiol. ~., 277-282.

WO 95/31544 63 PCT/fJS95105854 Goodwin, R.G., Anderson, D., Jerry, R., David, T., Brannan, C.L, Copeland, N.G., Jenkins, N.A.
and Smith, C.A. (1991) Mol. Cell Biol. ~ 3020-3026.
Gossen, M. and Boujard, H. (1992) Proc. Natl. Acad. Sci. USA, 89:547-5551.
Gryczan, T. (1982) The Molecular Bioloa~,of the Bacilli. Academic Press, N.Y.
307-329.
Guarente, L. (1983) in Methods Enrymol. ~Ol,, 181-191.
Harada, A., Sekido, N., Kuno, A., Akiyama, M., Kasahara, T., Nakanishi, L, Mukaid, and Matsushima, K. (1993) Int. Immunol. 5 681-690.
Heller, R.A. et al. (1990) Proc. Natl. Acad. Sci. USA, 87:6151-6155.
Hohmann, H.-P. et al. (1989) J. Biol. Chem., 264:14927-14934.
Holtmann, H. and Wallach, D. (1987) J. Immunol. 139. 1161-1167.
Itoh, N. et al. (1991) Cell 66:233.
Itoh, N. and Nagata, S. (1993) J. Biol. Chem. 268. 10932-7.
Izaki, K. (1978) Jpn. J. Bacteriol. 33, 729-742.
John, J.F. et al. (1986) Rev. Infect. Dis. 8 693-704.
1 ~ Joseph, S. and Burke, J.M. (1993) J. Biol. Chem. 268:24515-8.
Kendall, K.J. et al (1987) J. Bacteriol 1~9, 4177-4183. ' ,. _ , Khan, A.S. et al. (1992) Nature Genetics, 2_: 180-185.
Koizumi, M. et al. (1993) Biol. Pharm. Bull (Japan) 16 (9):879-83.
Kunkel, T.A. (1994) in : Current protocols in molecular biology, pp. 8.1.1-8.1.6 (Ausubel, F.M.
et al., eds.) Greene Publishing Associates, Inc. and Wiley & Sons, Inc., New York.
Loetscher, H., Pan, Y-C.E., Lahm, H.-W, Gentz, R., Brockhaud, M., Tabuchi, H.
and Lesslauer, W. (1990) Cell, 6:351-359.
Maniatis, T. et al. (1982) Molecular Cloning : A Labc~atorv Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor.
Maniatis, T. ( 1980) in : Cell Biology: A Comprehensive Treatise Vol. 3 : Gene Expression.
Academic Press, N.Y. 563-608.
Matsushima, K., Morishita, K., Yoshimura, T., Law, S., Kobayashi, Y., Lew, W., Appella, E., Kung, H.F., Leonard, E.J. and Oppenheim, J.J. (1988) J. Exp. Med. 167_ 1883-1893.
Nophar, Y. et al. (1990) EMBO J., x:3269-3278.
Oehm, A. et al. (1992) J. Biol. Chem. 267:10709.
Ogasawara, J., Watanabe-Fukunaga, R., Adachi, M., Matsuzawa, A., Kasugai, T., Kitamura, Y., Itoh, N., Suds, T. and Nagata, S. (1993) Nature 364. 806-809.
Okayama, H. (1983) Mol. Cell Biol. 3 280.
ONeal, K.D. and Yu-Lee, L.Y. (1993) Lymphokine Cytokine Res. ~2 309-312.
Piquet, P.F. t al. (1987) J. Exp. Med., C.~6:1280-89.
Realini, C., Ropers, S.W. and Rechsteiner, iii. (1994) FEBS Lett ~4$., 109-113.
Rechsteiner, M., Hoffman, L. and Dubiel, W. (1993) J. Biol. Chem. X68 6065-6068.
Sambrook et al. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring HarborLaboratory Press, Cold spring Harbor, NY.
Schall, T.J. et al. (1990) Cell, 61:361-370.

Schwalb et al. (1993) J. Bi :hem. 268 (14) :9949-54.
Seger, R. and Krebs, E.G. (1995) FASEB J. 9: 726-735.
Sekido, N., Mujaida, N., Harada, A., Nakanishi, L, Watanabe, Y., h4atsushima, K. (1993) Nature 365. 654-657.
Shimayama, T. et al., (1993) Nucleic Acids Symp. Ser. 29:177-8 .
Shore, S.K. et al. (1993) Oncogene 8_:3183-8.
Smith, C.A., David, T., Anderson, D., Solam, L., Beckmann, M.P., Jerzy, R.;
Dower,~wS.K., Cosman, D. and Goodwin, R.G. (1990) Science, 248:1019-1023. ~ - .
Smith, D.B. and Corcoran, L.M. (1994) in : Current protocols in molecular biology, pp: 16.7.1-16.7.8 {Ausubel, F.M. et al., eds.) Greene Publishing Associates, Inc. and Wiley & Sons, -Inc: New York.
Song, H.Y. et al. (1994) J. Biol. Chem. 2~6 , 22492-22495.
Stamenkovic, L, Clark, E.A. and Seed, B. (1989) Embo J. 8_:1403-1410.
Tartaglia, L. A., Ayes, T.M., Wong; G.H. and Goeddel, D.V. (1993) Cell, 74:845-853.
1 S Tracey, J.T. et al. (1987) Nature, 30:662-664: .
~Vallach, D. (1984) J. Immunol. _1~2 2464-9.
Wallach, D. (1986) in : Interferon 7 (Ion Gresser; ed.), pp. 83-122, Academic Press, London .
Wallach, D. et al. (1994) Cytokine 6 556.
Watanabe-Fukanaga; R., Brannan, C.L, Itoh, N., Yonehara, S., Copeland, N.G., Jenkins;~N:A:
and Nagata, S. (1992) J. Immunol. 148_ 1274-1279.
Watanabe-Fukunaga, R et al. (1992) Nature, 356_ 314-317.
Wiegmann, K., Schutze, S., Machleidt, T., Witte, D. and Kronke; M. (1994) Cell 78 1005=1015.
Wilks, A.F. et al. (1989) Proc. Natl. Acad. Sci. USA, ,8:1603-1607.
Zhao, J.J. and Pick, L. (1993) Nature (England) 365:448-51.

Claims (26)

1. A soluble, oligomeric tumor necrosis factor receptor (TNF-R) comprising at least two self-associated fusion proteins, each fusion protein having (a) at its one end, a TNF binding domain selected from the extracellular domain of a TNF-R, analogs or derivatives thereon said extracellular domain, analogs or derivatives thereof being incapable of deleterious self-association leading to interference of TNF binding or less than optimal TNF
binding, and being able to bind TNF; and (b) at its other end, a self-associating domain selected from (i) essentially all of the intracellular domain of the p55 TNF-R(p55-1C), extending from about amino acid residue 206 to about amino acid residue 426 of the native p55 TNF-R
molecule (p55-R); (ii) the death domain of the p55-IC extending from about amino acid residue 328 to about amino acid residue 426 of the native p55-R: (iii) essentially all of the intracellular domain of the Fas/APO1 receptor (Fas-IC); (iv) the death domain of Fas-IC; and (v) analogs, fractions or derivatives of any one of (i)-(iv) being capable of self association, wherein said at least two self-associated proteins self-associate only at said ends (b), having said ends (a) capable of binding to at least two TNF monomers, each end (a) capable of binding one TNF monomer; and salts and functional derivatives of said soluble, oligomeric TNF-R.

2. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) essentially all of the extracellular domain of the p55 TNF-R (p55-R) extending from about amino acid residue 1 to about amino acid residue 172 of the native p55-R, and as its at least two ends (b), essentially all of said p55-IC.

3. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) essentially all of the extracellular domain of the p55-R extending from about amino acid residue 1 to about amino acid residue 172 of the native p55-R and as its at least two ends (b) essentially all of said death domain of the p55-IC.

4. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) analogs or derivatives of the extracellular domain of the p55-R, each of said analogs or derivatives being capable of binding one TNF monomer, and being incapable of self-association, and as its at least two ends (b) essentially all of said p55-IC.

5. A soluble, oligomeric TNF-R according to claim 50 comprising as its two ends (a) analogs or derivatives of the extracellular domain of the p55-R, each of said analogs or derivatives being capable of binding one TNF monomer, and being incapable of self-association, and as its at least two ends (b) essentially all of said death domain of p55-IC.

6. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) essentially all of the extracellular domain of the p55-R extending from about amino acid residue 1 to about amino acid residue 172 of the native p55-R, and as its at least two ends, and as its at least two ends (b) essentially all of said Fas-IC.

7. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) essentially all of the extracellular domain of the p55-R extending from about amino acid residue 1 to about amino acid residue 172 of the native p55-R, and as its at least two ends, and as its at least two ends (b) essentially all of said death domain of Fas-IC.

8. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) analogs or derivatives of the extracellular domain of the p55-R, each of said analogs or derivatives being capable of binding one TNF monomer, and being incapable of self-association, and as its at least two ends (b) essentially all of said Fas-IC.

9. A soluble, oligomeric TNF-R according to claim 50 comprising as its at least two ends (a) analogs or derivatives of the extracellular domain of the p55-R each of said analogs or derivatives being capable of binding one TNF monomer, and being incapable of self-association, and as its at least two ends (b) essentially all of said death domain of Fas-IC.

10. A process for the production of the soluble, oligomeric TNF-R according to any one of claims 50-58 comprising:
(a) the construction of an expression vector encoding any one of said fusion proteins, the DNA sequence of each of said ends of the fusion protein being obtained from cloned DNA
sequences encoding essentially all of said extracellular domain of the TNF-R, analogs or derivatives thereof; and from cloned DNA sequences encoding essentially all of said p55-IC, p55-IC death domain, Fas-IC, Fas-IC death domain, analogs or derivatives of all of the aforegoing, said ends being ligated together to form a fusion protein sequence, and said fusion protein sequence being inserted into said vector under the control of transcriptional and translational regulatory sequences;
(b) introduction of the vector of (a) into a suitable host cell in which said fusion protein is expressed; and (c) purification of the fusion protein expressed in said host cells, said fusion protein self-associating prior to, during, or following the purification process to yield a soluble, oligomeric TNF-R.

11. An expression vector comprising a fusion protein sequence encoding said fusion proteins of any one of claims 50-58.

12. A vector according to claim 60 for use in a process according to claim 59.

13. A host cell containing a vector according to claim 60 capable of expressing said fusion protein sequence.

14. A pharmaceutical composition comprising the soluble, oligomeric TNF-R, salts or functional derivatives thereof and mixtures of any of the foregoing, according to any one of claims 50-58, as active ingredient together with a pharmaceutically acceptable carrier.

15. A soluble, oligomeric TNF-R, salts or functional derivatives thereof and mixtures of any of the foregoing, according to any one of claims 50-58, for use in antagonizing the deleterious effect of TNF in mammals, especially in the treatment of conditions wherein an excess of TNF is formed endogenously or is exogenously administered.

16. A soluble, oligomeric TNF-R, salts or functional derivatives thereof and mixtures of any of the foregoing according to any one of claims 50-58, for use in maintaining prolonged beneficial effects of TNF in mammals, when used with TNF exogenously administered.

17. A soluble, oligomeric Fas/APO1 receptor (Fas-R) comprising at least two self-associated fusion proteins, each fusion protein having (a) at its one end, a Fas ligand binding domain selected from the extracellular domain of a Fas-R, analogs or derivatives thereof being incapable of self-associating and being able to bind Fas ligand; and (b) at its other end, a self-associating domain selected from (i) essentially all of the intracellular domain of the p55 TNF-R (p55-IC), extending from about amino acid residue 206 to about amino acid residue 426 of the native p55 TNF-R molecule (p55-R); (ii) the death domain of the p55-IC
extending from about amino acid residue 328 to about amino acid residue 426 of the native p55-R; (iii) essentially all of the intracellular domain of the Fas/APO1 receptor (Fas-IC);
(iv) the death domain of Fas-IC; and (v) analogs or derivatives of any one of (i)-(iv) being capable of self association, wherein said at least two self associated proteins only self associate at said ends (b) having said ends (a) capable of binding to at least two Fas ligand monomers, each end (a) capable of binding one Fas ligand monomer; and salts and functional derivatives of said soluble, oligomeric Fas-R.

18. A process for the production of the soluble, oligomeric Fas-R according to claim 66 comprising:
(a) the construction of an expression vector encoding any one of said fusion proteins, the DNA sequence of each of said ends of the fusion protein being obtained from cloned DNA
sequences encoding essentially all of said extracellular domain of the Fas-R, analogs or derivatives thereof ; and from cloned DNA sequences encoding essentially all of said p55-IC, p55-IC death domain, Fas-IC, Fas-IC death domain, analogs or derivatives thereof of all the aforegoing, said ends being ligated together to form a fusion protein sequence, and said fusion protein sequence being inserted into said vector under the control of transcriptional and translational regulatory sequences;
(b) introduction of the vector of (a) into a suitable host cell in which said fusion protein is expressed; and (e) purification of the fusion protein expressed in the host cells, said fusion protein self-associating prior to during, or following the purification process to yield a soluble, oligomeric Fas-R.

19. An expression vector comprising a fusion protein sequence encoding said fusion proteins of claim 66.

20. A vector according to claim 68 for use in a process according to claim 67.

21. A host cell containing a vector according to claim 68 capable of expressing said fusion protein sequence.

22. A pharmaceutical composition comprising the soluble, oligomeric Fas-R, salts or functional derivatives thereof and mixtures of any of the foregoing, according to claim 66 as active ingredient together with a pharmaceutically acceptable carrier.

23. A soluble, oligomeric Fas-R, salts or functional derivatives thereof and mixtures of any of the foregoing, according to claim 66, for use in antagonizing the deleterious effect of Fas ligand in mammals, in the treatment of conditions wherein an excess of Fas ligand is formed endogenously or is exogenously administered.

24. A soluble, oligomeric receptor having affinity for both TNF and FAS-R
ligand (mixed affinity receptor), comprising at least two self-associated fusion proteins, one of which fusion proteins is a TNF-specific TNF-R-derived protein of any one of claims 50-58; and the other fusion protein is a FAS-R ligand-specific FAS-R-derived protein of claim 66.

25. A pharmaceutical composition comprising the mixed amity receptor according to claim 73.

26. A mixed affinity receptor according to claim 73 for use in antagonizing the deleterious effects of TNF and FAS-R ligand in mammals.