WO2002034780A2 - Hetero-oligomeric compounds derived from the tnf cytokine superfamily and their use - Google Patents

Abstract

The invention relates to a method to form artificial, hetero-oligomeric complexes of proteins that do normally occur in homo-oligomeric complexes. Said hetero-oligomeric complexes do have a biological activity that is similar or identical to the activity of the homo-oligomeric complexes.The current invention also relates to a method to use said hetero-oligomeric complexes to screen for compounds that can stabilize said hetero-oligomeric complexes, or to screen for compounds that dissociate said hetero-oligomeric complexes into monomers, which are not biologically active, or show a biological activity, which is clearly distinct from that of the hetero-oligomeric complexes, or to screen for compounds that inhibit the oligomerisation or reoligomerisation of monomers in oligomeric complexes.Furthermore, the invention relates to pharmaceutical compositions comprising said compounds.

Description

Methods and means for the deoligomerisation of oligomeric molecules

The invention relates to a method to form artificial, hetero-oligomeric complexes of proteins that do normally occur in homo-oligomeric complexes. Said hetero-oligomeric complexes do have a biological activity that is similar or identical to the activity of the homo-oligomeric complexes.

The current invention also relates to a method to use said hetero-oligomeric complexes to screen for compounds that can stabilize said hetero-oligomeric complexes, or to screen for compounds that dissociate said hetero-oligomeric complexes into monomers, which are not biologically active, or show a biological activity, which is clearly distinct from that of the hetero-oligomeric complexes, or to screen for compounds that inhibit the oligomerisation or reoligomerisation of monomers in oligomeric complexes. Furthermore, the invention relates to pharmaceutical compositions comprising said compounds.

The excessive and/or inappropriate production of growth factors, lymphokines and cytokines plays an important role in several diseases. Although the cytokine mediated response is an essential part of the normal response to trauma and infection, excessive production of pro-inflammatory cytokines or production of cytokines in a wrong biological context is associated with a wide range of diseases, such as sepsis, rheumatoid arthritis, inflammatory bowel disease, cancer, response to malaria and AIDS. Several of these growth factors, cytokines and cytokine-like molecules are oligomeric molecules in their biologically active form. As a non-limiting example, Placenta Growth Factor (PIGF) and Vascular Endothelial Growth Factor (VEGF) are forming dimeric complexes. Members of the Tumour Necrosis Factor (TNF) superfamily do form trimeric complexes. Up to now, the latter family comprises 19 known members (LTA, TNF, LTB, PGL.YRP, TNFSF4, TNFSF5, TNFSF6, TNFSF7, TNFSF8, TNFSF9, TNFSF10, TNFSF11 , TNFSF12, TNFSF13, TNFSF13B, TNFSF14, TNFSF15, TNFSF18 (all names according to the Human Genome nomenclature database) and TNFSF20 (Tribouley et al., 1999). With the exception of LTA, which is the only entirely secreted protein, members of this superfamily are transmembrane proteins that can however be proteolytically processed and act as a soluble cytokine. The extracellular domain has β jelly roll structure (Banner et ai, 1993) and is important in ligand trimerisation. Intrinsic to this trimerisation is the formation of a receptor-binding site at the junction between neighbouring subunits, creating a multivalent ligand. As a consequence, ligands of the TNF superfamily are only biologically active as homotrimeric or natural heterotrimeric protein complexes and not as monomeric protein. Although all members of the TNF cytokine family seem to be critically involved in the regulation of infections, inflammation, autoimmune diseases and/or tissue homeostasis (Smith et ai, 1994), TNF is the best studied member of this family. Originally, TNF was identified as a systemic mediator of endotoxemic shock, cachexia, and tumour regression. TNF is a potent cytokine that exerts a wide range of activities in inflammatory and immune reactions and is also an important mediator in many diseases. It exists both as a transmembrane and as a secreted, homotrimeric complex that signals via interaction with two types of the TNF receptor, TNF-R1 (p55 TNF-R; CD120a) and TNF-R2 (p75 TNF-R; CD120b), One TNF homotrimeric protein complex can bind three TNF-receptor molecules. Binding of the ligand to the receptors induces a clustering of the intracellular domains, initiating the intracellular signalling cascades. As the formation of a homotrimeric protein complex is essential for the receptor binding, dissociation of the homotrimeric complex into monomeric TNF impairs its biological activity.

Treatment with some reagents such as Triton X-100, DMSO or guanidinium chloride promotes the dissociation of TNF. This dissociation can be reversed by enhancing the TNF concentration or removing the chemical agent (Hlodan and Pain, 1995; Corti et ai., 1992; Hlodan and Pain, 1995). An excess of soluble TNF- receptors, which only bind to trimeric TNF, on the other hand can stabilise the biological activity of subnanomolar amounts of TNF during incubation, presumably by diminishing the dissociation of TNF (Aderka et ai, 1992).

As growth factors, lymphokines and cytokines in general and TNF in particular are involved in the pathogenesis of several inflammatory diseases, they are obvious and attractive therapeutical targets. Several approaches have been undertaken in order to inhibit the deleterious biological effects of TNF. Experimental studies have shown that TNF inhibition by administration of anti-TNF monoclonal antibodies or soluble TNF receptors (or TNF receptor fusion proteins) is effective in the treatment of rheumatoid arthritis and inflammatory bowel disease (Camussi and Lupia, 1998; Sandborn and Hanauer, 1999).

Isolation of alternative compounds that can either inactivate or stabilise growth factors, lymphokines and cytokines, and particularly TNF is therefore an important aim in pharmaceutical research. The activity of growth factors, lymphokines and cytokines is often measured in biological assays. However, these assays are not suitable to screen for inhibitors, because it is unclear at which level the inhibitor is operating. Indeed, if in such an assay inhibition is detected, it can be due to an inactivation of the ligand, as well as to an inactivation of the receptor or of one of the compounds of the signalling pathway. Compounds that act on the latter levels may cause unwanted side effects. In cases where the monomeric subunit is biologically inactive or has a biological activity that is clearly distinct form the oligomer, a screening method for compounds that dissociate or stabilise the oligomeric protein complex is an attractive alternative. At present, the dissociation/reassociation of oligomeric protein complexes such as PIGF, VEGF or TNF is studied by several methods, which are, however, not suited for application in large-scale screening of chemical reagents. Size exclusion chromatography of labeled TNF is a reliable but labour-intensive technique. The difference in intrinsic tryptophan fluorescence (excitation wavelength at 280 nm, emission at 320 nm) between trimeric and monomeric TNF is also labour-intensive and prone to interference by light absorption of the tested compounds (Hlodan and Pain, 1995). ELISA's and RIA's with anti-TNF monoclonal antibodies do not show a strict specificity for trimeric and monomeric forms, possibly due to the fact that the antibody used can influence the equilibrium between trimeric and monomeric forms of TNF (Corti et ai, 1992). The present invention describes a method to screen for compounds that can stabilise or destabilise a homo-oligomeric protein complex, by the use of a hetero-oligomeric protein complex, comprising one or more modifications of the subunit of said homo- oligomeric protein complex, whereby said hetero-oligomeric protein complex is preferentially retaining essentially the same biological activity as said homo-oligomeric protein complex. Each modified subunit can be distinguished from the non-modified subunit and/or from those subunits carrying another modification. The hetero- oligomeric complex can consist of two or more different kinds of modified subunits without any unmodified subunit, or it can consist of one or more different kinds of modified subunits, with one or more unmodified subunits. As long as the hetero- oligomeric protein complex is stable, selective capturing of either a modified or a non- modified subunit, e.g. by immunoprecipitation or binding on a solid substrate, will result in a capturing of the whole protein complex. However, when a destabilising compound is added, selective capturing will only capture the monomeric subunit to which it is directed and the remaining subunits will be released. The effect can be measured by measuring either the amount of units that are released, or the amount of subunits that are retained after a certain time period. As normally most of the oligomeric complexes do show a spontaneous destabilisation with time, not only destabilising compounds, which enhance the destabilisation, can be screened, but also compounds that do stabilise the complex and slow down the disaggregation.

One aspect of the invention is an artificial, biologically active hetero-oligomeric protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit can form a biologically active homo-oligomeric protein complex on its own, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero-oligomeric protein complex, and whereby said unmodified subunit as well as said modifications have a clearly distinct biological activity as monomeric molecule, compared to the oligomeric complexes. Preferentially, the monomeric molecules are biologically inactive. Preferably, said hetero-oligomeric and/or said homo-oligomeric protein complex is not a receptor and/or the oligomeric complex is not membrane bound. More preferably, said hetero-oligomeric and/or said homo-oligomeric protein complex is a ligand. Preferably said hetero-oligomeric protein complex and said homo-oligomeric protein complex are composed of the same number of subunits. More preferably, both protein complexes are trimeric. Said modifications should be easily distinguishable from the non-modified form, or, if more than one modified form is used, from each other. As non-limiting examples, modifications can be¹²⁵l iodinated forms, biotinylated forms, subunits carrying a tag such as a his-tag or subunits, comprising one or more amino acid changes, preferentially resulting in the creating or destruction of a specific epitope. Homologous molecules, showing at least 40% identities and/or 65% positives at protein level, as measured by BL-ASTP (Altschul et al., 1997), are still considered as modifications of the same subunit. Preferably, said homologous molecules show at least 60% identities, more preferably at least 75% identities, most preferably 80% identities. A preferred embodiment is a hetero-oligomeric protein complex whereby said subunit is a subunit of a member of the TNF superfamily. Another preferred embodiment is a hetero-oligomeric protein complex whereby said subunit is a TNF subunit. A most preferred embodiment is a hetero-oligomeric complex whereby said unmodified subunit is a human TNF subunit, and said modification a murine TNF subunit, or said unmodified subunit is a murine TNF subunit, and said modification a human TNF subunit.

Another aspect of the invention is the use of a biologically active hetero-oligomeric protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit can form a biologically active homo-oligomeric protein complex on its own, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero-oligomeric protein complex, and whereby said unmodified subunit as well as said modifications have a clearly distinct biological activity as monomeric molecule, compared to the oligomeric complexes, to screen compounds that stabilise or destabilise said hetero-oligomeric protein complex. Preferably, said hetero-oligomeric protein complex is an artificial hetero-oligomeric protein complex. Even more preferably said hetero-oligomeric complex comprises a human TNF subunit as unmodified subunit, and a murine TNF subunit as modification, or a murine TNF subunit as unmodified subunit, and a human TNF subunit as modification.

Still another aspect of the invention is a method to screen compounds that stabilize or destabilize a biologically active hetero-oligomeric protein complex said protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit can form a biologically active homo-oligomeric protein complex on its own, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero-oligomeric protein complex, and whereby said unmodified subunit as well as said modifications have a clearly distinct biological activity as monomeric molecule, compared to the oligomeric complexes, said method comprising a) binding one or more of the subunits in a specific way; b) contacting said hetero-oligomeric protein complex with one or more of said compounds; c) measuring the amount of one or more subunits present in monomeric and/or oligomeric form. The binding can be covalently or not, and can be realized before the formation of the hetero-oligomeric complex, e.g. by binding a monomeric subunit to a solid substrate, followed by oligomerisation by contacting the immobilized subunit to free subunits, or after formation of the hetero-oligomeric complex, e.g. by specific binding of the non-modified subunit or one of the modifications by an antibody. Contacting said hetero-oligomeric protein complex with one or more of said compounds may be realized before or after said binding. It is indeed obvious for the person skilled in the art that the stability can be measured, as non-limiting examples, by measuring the release of subunits from an immobilized hetero-oligomeric complex, as well as by measuring the amount of said hetero-oligomeric complex by precipitation of said complex, e.g. by immunoprecipitation, after a certain contact time with said stabilising or destabilising compound. Quantification of monomers of oligomers can be realized, as non-limiting examples, by radioactivity or by immunological techniques. Preferably, said hetero-oligomeric protein complex is artificial. A preferred embodiment is a method according to the invention whereby said hetero- oligomeric complex comprises a human TNF subunit as unmodified subunit, and a murine TNF subunit as modification, or a murine TNF subunit as unmodified subunit, and a human TNF subunit as modification.

Still another aspect of the invention is a compound, isolated with the use of a hetero- oligomeric protein complex according to the invention. Still another aspect of the invention is a pharmaceutical preparation, comprising a compound according to the invention.

Definitions

The following definitions are set forth to illustrate and define the meaning and scope of various terms used to describe the invention herein.

Hetero-oligomeric protein complex as used here means a protein complex that consists of one or more identical subunits (modified or not) with at least one other subunit (modified or not), whereby latter subunit is clearly distinct from the former. Artificial hetero-oligomeric complex means that the complex doesn't exist as such in nature. As a non-limiting example, said artificial hetero-oligomeric complex may comprise subunits coming from different sources such as different organisms, or it may comprise subunits carrying induced mutations, or it may comprise chemical modifications of the protein. Comprising one or more modifications of one subunit means that the protein complex may comprise one or several modified forms of the same subunit, whereby the modification may be identical or different. •

Subunit as used here can be either an unmodified or a modified subunit. - Modification as used here may be any modification, as long as the modification allows a selective distinction of the modified form from the other forms of the subunit, modified by another modification or not modified. Modification can be, as a non limiting example, radioactive labelling such as iodination, biotinylation, or the change of one or more amino acids preferentially leading to the loss or the generation of a specific epitope. The replacement of a subunit by a homologous subunit of the same or another species is still considered as being a modification, as long as the homology is at least 40% identities and/or 60% positives, preferably 60% identities, more preferably

75% identities, most preferably 80% as measured at protein level by BLASTP (Altschul et al., 1997). Covalent binding of subunit to a solid support is also considered a modification of this subunit.

Biologically active hetero-oligomeric protein complex refers to the biological activity of the homo-oligomeric protein complex within the cell type and/or organism in which it is normally functional. In the special case of TNF, biologically active means that the trimeric complex can bind to the TNF receptors, as can be measured as described below.

Can form a biologically active homo-oligomeric protein complex means that the unmodified subunit, in absence of modified subunits, when brought in a suitable medium will form spontaneously a homo-oligomeric complex on its own, without interference of another protein, whereby said homo-oligomeric complex has a biological activity that is clearly distinct of that of the monomer.

Essentially identical as used here means that there is no essential change in function between the hetero-oligomeric protein complex and the homo-oligomeric protein complex within the cell type and/or organism in which the homo-oligomeric protein complex is tested, but does not exclude a change in specific activity

Brief description of the figures

Figure 1 : Gel filtration experiment using TNF homo- and heterotrimers, in absence of destabilising compounds. TNF deoligomerises during incubation at low concentration and monomers as well as trimers can be detected by gelfiltration (muTNF at < 400 pM ; huTNF at < 100 pM ). Deoligomerisation of labeled muTNF can be prevented by addition of excess, unlabeled muTNF or huTNF.

Figure 2: Effect of TNF concentration on the trapping of labeled, homotrimeric or heterotrimeric TNF in a RIA using the anti-human TNF antibody 61 E71 or the anti- murine TNF antibody 1 F3F3.¹²⁵l-human TNF was preincubated with different concentrations of unlabeled murine TNF (squares) or human TNF (circles). The mixture was applied to wells coated with 61 E71 , an anti-human TNF antibody (black symbols), or coated with 1 F3F3, an anti-murine TNF antibody (white symbols). Wells were washed and bound label was counted and is represented as percentage of label applied to well.

Figure 3: Binding of labeled huTNF/murine TNF heterotrimers on soluble murine TNF- receptors.¹²⁵l-human TNF was incubated with unlabeled murine TNF (black symbols) or human TNF (white symbols) for 0 (circles) or 16 hours (triangles). The mixture was applied to wells coated with soluble murine TNF-R1 (panel A) or coated with soluble murine TNF-R2 (panel B). Bound label was measured in a gamma-counter.

Figure 4: Binding of labeled murine TNF-75/murine TNF heterotrimers on soluble murine TNF-receptors.¹²⁵l-murine TNF-75 was incubated with unlabeled murine TNF for 24 hours and subsequently applied to wells coated with soluble murine TNF-R1 (white circles) or coated with soluble murine TNF-R2 (black triangles). Bound label was measured.

Figure 5: Screening assay with labeled human TNF/murine TNF heterotrimers in a RIA using the anti-murine TNF antibody 1 F3F3.¹²⁵l-human TNF was preincubated with unlabeled murine TNF and the mixture was applied to wells coated with the anti- murine TNF antibody 1 F3F3 to trap heterotrimeric TNF. After washing away unbound label, the wells were treated with different agents, as indicated in the legend. After 15 minutes, 4 or 18 hours incubation, the wells were washed and remaining label was counted and was represented as percentage of label bound at start of treatment. Figure 6: Screening assay with labeled human TNF/murine TNF heterotrimers in a RIA using the anti-murine TNF antibody 1F3F3.¹²⁵l-human TNF was preincubated with unlabeled murine TNF and the mixture was applied to wells coated with the anti- murine TNF antibody 1 F3F3 to trap heterotrimeric TNF. After washing away unbound label, the wells were treated with different agents, as indicated in the legend. After 15 minutes, 4 or 18 hours incubation, the wells were washed and remaining label was counted and was represented as percentage of label bound at start of treatment. Figure 7: Screening assay with labeled murine TNF/human TNF heterotrimers in a RIA using the anti-human TNF antibody 61 E71.¹²⁵l-murine TNF was incubated with unlabeled human TNF and the mixture was ' applied to wells coated with antibody 61 E71. After washing away unbound label, the wells were treated with different agents for 15 minutes, 4 or 18 hours. After the treatment, wells were washed and the label present in the wells and in the wash-solutions was counted. The amount of label in the two fractions is represented as percentage of total label in both fractions. Figure 8: Effect of TNF concentration on the formation and trapping of heterotrimeric TNF on streptavidin-coated wells.¹²⁵l-human TNF was preincubated with different concentrations of biotinylated human TNF and applied to wells coated with streptavidin. The wells were washed after 2 hours, and bound label was counted and is represented as percentage of label applied to well.

Figure 9: Screening assay using labeled human TNF/biotinylated human TNF.¹²⁵l-human TNF was preincubated with biotinylated human TNF and the mixture was applied to streptavidin-coated wells. After washing away unbound label, the wells were treated with different agents. After 15 minutes, 4 or 18 hours incubation, the wells were washed and the remaining label was measured and was represented as percentage of label that was bound at start of treatment. Figure 10: Gel filtration of labeled human TNF after pretreatment with different agents. Labeled human TNF was incubated with different agents, each at a concentration of 100 μg/ml, for 1 hour at 22°C. The mixture was chromatographed on a Sephacryl S100 column and the radioactivity present in each fraction was measured and was represented as percentage of label present in all fractions. In this chromatogram three peaks can be observed; a peak with trimeric TNF (fractions 20-36), one with monomeric TNF (fractions 36-44) and one with degradation products (fractions 54-80). Figure 11: Treatment of murine TNF heterotrimers, trapped on ExtrAvidin. l-murine TNF was preincubated with biotinylated murine TNF and the mixture was applied to ExtrAvidin-coated wells. After washing away unbound label, the wells were treated with different agents. After 0, 2, 4 or 18 hours incubation, bound label was measured and this was represented as percentage of label, bound at start of treatment.

Figure 12: Gel filtration of murine TNF. Labeled murine TNF was preincubated with different concentrations of unlabeled murine TNF (total TNF concentration is indicated in legend) as such or in the presence of methylene blue (MB). The mixture was chromatographed on a Sephacryl S100 column and the radioactivity in each fraction was measured and is here represented as percentage of label present in all fractions.

Examples Materials and Methods to the examples

Cytokines and antisera

Purified Escherichia coli-derived muTNF, huTNF and huLT were produced in our laboratory and had a specific biological activity of 2.1 x 10⁸ U/mg, 8.4 x 10⁷ U/mg, and 4.0 x 10⁷ U/mg respectively in the L929 cytotoxicity assay. Mutein muTNF75 was created by site-specific mutagenesis of amino acid Arg 32 to Tyr and Ala 145 to Arg. This mutein binds only to muTNF-R2 in an in vitro receptor-binding assay. It is not cytotoxic towards L929. Monoclonal antibody to muTNF, 1 F3F3 (Lucas et al., 1990), was a kind gift from Dr. R. Lucas and Dr. P. De Baetselier (Free University of Brussels, Belgium). Monoclonal antibodies to huTNF, 61 E71 (Stevens et al., 1990), and to muTNF, TN3 (Sheehan et al., 1989) were kindly provided by Dr. W. Buurman (University of Limburg, The Netherlands).

¹²⁶1 radiolabeling ofligands

E. coli-derived muTNF and huTNF were radiolabeled using the lodogen iodination agent according to the manufacturer's instructions (Pierce Chemicals, Rockford, IL). The labeled proteins were separated from unincorporated radioactivity on a G-25 column (PD10, Pharmacia-LKB Biotechnology, Uppsala, Sweden) and had a specific radioactivity of 10-50 μCi/μg.

Biotinylation of ligands

E. coli-derived muTNF and huTNF were biotinylated using a NHS-LC-biotinylation kit, according to the manufacturer's instructions (Pierce Chemicals, Rockford, IL). The biotinylated proteins were separated from unincorporated NHS-LC-biotin on a G-25 column (PD10, Pharmacia-LKB Biotechnology, Uppsala, Sweden).

Gel-filtration

Gel-filtration was performed on a Sephacryl S-100 column (Pharmacia-LKB Biotechnology, Uppsala, Sweden). The column was equilibrated and eluted in phosphate buffered saline containing 0.02% BSA and 0.02% NaN3 (PBS/BSA) at a flow rate of 0.4 ml/min. Fractions of 0.3 ml were collected and tested for radioactivity in a gamma counter. All gel-filtration chomatographies were performed at 4°C.

Radioimmunoassay (RIA) 96-well plates (Maxi Breakapart, Nunc) were incubated overnight at 4°C with monoclonal antibody (2.5 μg/ml; 100 μl/well) in PBS. After coating, wells were saturated for 3 h at 26°C with 150 μl/well of PBS containing 2% BSA. After washing with PBS, plates were incubated with samples containing radiolabeled TNF, diluted in PBS/BSA, (100 μl/well) for 90 min at 26°C. After washing with PBS, wells were separated and individually counted in a gamma counter.

Receptor binding assay

The soluble (extracellular) fragment of murine TNF-R1 (smuTNF-R1) or smuTNF-R2 were produced with the baculovirus-expression system in Sf9 insect cells and partially purified. Microtiter plates (Nunc-lmmuno Breakapart Module) were coated with smuTNF-R1 (0.66 μg/ml) or smuTNF-R2 (1 μg/ml) overnight at 4°C. Blocking was with PBS/2% BSA. The microtiter plates were then incubated with 400 pM (20 ng/ml)¹²⁵l- labeled TNF in the presence of different concentrations of unlabeled TNF. After 4 hours incubation at 26°C, the wells were washed with PBS/ 0.02% BSA and bound radioactivity was measured in a gamma-counter.

L929 cytotoxicity assay

L929 murine fibrosarcoma cells (Rega Institute Leuven, Belgium) were grown in

Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum, penicillin (100 U/ml), streptomycin sulphate (100 μg/ml) and L-glutamine (2 mM) (DMEM/FCS). Cells were seeded in 96-well microtiter plates at 30000 cells/well. The next day, a serial dilution of TNF in DMEM/FCS with or without mutein muTNF75 was added to the cells in the presence of actinomycin D (1 μg/ml). After 18 h incubation, the amount of surviving cells was determined by the MTT colorimetric method (Mosmann, 1983). Example 1: TNF heterotrimeric complexes are formed during coincubation of different TNFs

Murine TNF was labeled with¹²⁵l using IODO-GEN iodination reagent (Pierce, Rockford, USA). Labeled murine TNF was subjected to size exclusion chromatography on a Sephacryl S-100 column where after the radioactivity in the different fractions was measured in a Gamma 5500B counter (Beckman, USA). A significant part of the labeled murine TNF was found to dissociate into monomers after overnight incubation at the subnanomolar range (20 ng/ml [= 400 pM] or less), at 25°C, even in the complete absence of denaturing agents. When labeled murine TNF (400 pM) had been incubated overnight in the presence of unlabeled murine TNF (4 nM), almost all label remained in the fractions containing trimeric TNF. Surprisingly, the amount of labeled TNF in the trimeric fraction was not only enhanced by coincubation with murine TNF but also by coincubation with human TNF, which is 78% homologous to the murine TNF at the amino acid level (Figure 1). 61 E71 , a monoclonal antibody specific for human TNF (Dr. W. Buurman, University of Limburg, The Netherlands) did not retain labeled murine TNF in a RIA assay, as expected. However after prior incubation of labeled murine TNF (200 pM) with human TNF (1 nM) for 16 hours at 25°C, almost 15% of total label was bound by 61 E71. In a similar way, 1 F3F3, a monoclonal antibody specific for murine TNF (R&D Systems), did not bind labeled human TNF as such, but 18%) of total label was retained after prior incubation with murine TNF, proving that heterotrimeric protein complexes between murine and human TNF are formed.

Radiolabeled human TNF (20 ng/ml) was coincubated with different concentrations of unlabeled murine TNF or human TNF at 26°C for 24 hours to form heterotrimeric as well as homotrimeric TNF. The preincubated mixture was then allowed to bind on wells of 96 well plates (Nunc-lmmuno Breakapart, Nunc), coated with monoclonal antibody against murine TNF (1 F3F3, R&D Systems) or with monoclonal antibody against human TNF (61 E71 , Dr. W. Buurman, University of Limburg, The Netherlands). After 90 minutes, the wells were washed with PBS/BSA (0.02%) and the label bound on the wells was measured in a Gamma 5500B counter (Beckman, USA) (Figure 2). This experiment indicates how the optimal TNF concentrations to trap previously formed TNF heterotrimers on coated wells, can be measured. Example 2: Heterotrimeric TNF molecules still interact with TNF-receptors: In vitro receptor binding

Radiolabeled human TNF (20 ng/ml) was coincubated with different concentrations of unlabeled murine TNF or human TNF at 26°C for 0 or 16 hours and subsequently allowed to bind on wells from a microtiter plate (Nunc-lmmuno Breakapart, Nunc), coated with soluble murine TNF-R1 or with soluble murine TNF-R2. After 4 hours incubation, the wells were washed with PBS/BSA (0.02%) and bound label was measured in a gamma-counter (Figure 3). As expected on the base of the affinity of the receptors, labeled human TNF only binds to the soluble murine TNF-R1 (Tartaglia et al., 1991). After prior coincubation with unlabeled murine TNF, however, a significant amount of label, now present in heterotrimeric molecules, is retained on wells coated with soluble murine TNF-R2.

Receptor-specific muteins of human TNF have been made and described (Van Ostade et al., 1993). We have created a similar mutein of murine TNF that interacts specifically with murine TNF-R2. This mutein, murine TNF-75, was created by site- specific mutagenesis of murine TNF resulting in the replacement of amino acids Arg 32 with Tyr and Ala 145 with Arg. Radiolabeled murine TNF-75 (20 ng/ml) was coincubated with different concentrations of unlabeled murine TNF at 26°C for 24 hours and subsequently allowed to bind on wells from a microtiter plate (Nunc-lmmuno Breakapart, Nunc), coated with soluble murine TNF-R1 or with soluble murine TNF-R2. After 4 hours incubation, the wells were washed with PBS/BSA (0.02%) and bound label was measured in a gamma-counter (Figure 4). Whereas labeled murine TNF-75 only binds to soluble murine TNF-R2, part of the label was also retained on wells coated with soluble murine TNF-R1 after prior coincubation of labeled murine TNF-75 with unlabeled murine TNF.

Example 3: Heterotrimeric TNF molecules still interact with TNF-receptors: In vivo activity

Murine TNF was serially diluted in DMEM/FCS, or in DMEM/FCS containing murine TNF-75 (200 pM or 600 pM), and directly or after a preincubation of 24 hours at 37°C, 100 μl of the dilution samples was added to wells containing L929s fibrosarcoma cells (30000 cells/well). After incubation (18 hours at 37°C) in the presence of actinomycine D (1 μg/ml), surviving cells were measured by MTT staining. The biological activity of the samples is represented in U/ml (1 U/ml is the concentration of TNF sufficient to kill half of the cells) (Table 1). The TNF-R2-specific murine TNF-75 is not cytotoxic to L929 cells at the concentrations used. The apparent biological activity of murine TNF is strongly decreased upon preincubation in DMEM/FCS, due to dissociation into inactive monomers. However, when murine TNF-75 was present during the preincubation of murine TNF, the higher concentration of TNF monomers was promoting trimer formation and the decrease of murine TNF activity was reduced due to the formation of heterotrimeric, functional TNF. From the results, it is clear that heterotrimeric molecules containing two wild type TNF subunits are still able to bind to cell bound murine TNF-R1 and to induce the intracellular signalling cascade leading to L929s cytotoxicity.

Example 4: Development of a screening assay using human TNF/mouse TNF heterotrimeric protein complexes

Radiolabeled human TNF (20 ng/ml) was coincubated with unlabeled murine TNF (50 ng/ml) at 22°C for 16 hours to form heterotrimeric as well as homotrimeric TNF. The preincubated mixture (100 μl/well) was then allowed to bind on wells of 96 well plates (Nunc-lmmuno Breakapart, Nunc), coated with monoclonal antibody against murine TNF (1F3F3, R&D Systems). After 90 minutes, the wells were washed with PBS/BSA (0.02%)). Hereby unlabeled murine TNF and labeled heterotrimeric TNF were retained. Subsequently, the wells were treated with different concentrations of Triton X-100, DMSO, guadinium chloride or methylene blue (Oterop generics), and after incubation for 15 minutes, 4 h or 18 h, washed with PBS/BSA. The radioactivity remaining in the individual wells was counted in a Gamma 5500B counter. Alternatively the radioactivity during the treatment and the subsequent washes can be counted. The results are summarised in Figure 5.

The same set up was used to test the destabilising effect of Triton X-100, methylene blue, methylene green (Sigma, M-7766), basic blue 24 (Sigma, B-4631) and brilliant cresyl blue (Sigma, B-2002). The results are summarised in figure 6. In a similar way, labeled murine TNF (60 ng/ml) was coincubated with unlabeled human TNF (200 ng/ml). This mixture was applied to 96 well plates coated with anti- human TNF antibody (61E71 , Dr. W. Buurman, University of Limburg, The Netherlands). Subsequently the wells were treated with Triton X-100, DMSO, methylene blue, or with monoclonal antibodies against trimeric murine TNF, (1 F3F3; R&D Systems) or monomeric murine TNF (TN3; Dr. W. Buurman, University of Limburg, The Netherlands). The radioactivity remaining in the individual wells as well as the radioactivity released during the treatment and the subsequent washes, was counted in a Gamma 5500B counter. The results are summarised in Figure 7.

Example 5: Development of a screening assay using biotinylated human TNF/iodinated human TNF heterotrimeric protein complexes

Biotinylated human TNF was prepared using a NHS-LC-biotinylation kit (Pierce, USA) and human TNF was labeled with¹²⁵l using IODO-GEN iodination reagent (Pierce, Rockford, USA). Radiolabeled human TNF (100 ng/ml) was coincubated with different concentrations of biotinylated human TNF at 26°C for 16 hours to form heterotrimeric TNF, containing as well¹²⁵l-labeled as biotinylated TNF subunits. The preincubated mixture was then allowed to bind on wells of microtiter plates (Nunc-lmmuno Breakapart, Nunc), coated with different concentrations of streptavidin (Sigma, USA). After 120 minutes, the wells were washed with PBS/BSA (0.02%) and the label bound on the wells was measured in a Gamma 5500B counter (Beckman, USA) (Figure 8). This experiment illustrates how the optimal TNF concentrations to form and trap biotinylated TNF heterotrimers on streptavidin-coated wells, can be measured. Radiolabeled human TNF (70 ng/ml) was incubated with biotinylated human TNF (300 ng/ml) for 16 hours at 26°C and the mixture was allowed to bind on wells coated with streptavidin (0.25 μg/ml). After 120 minutes, the wells were washed with PBS/BSA (0.02%>), hereby retaining biotinylated homotrimeric (unlabeled) and heterotrimeric (labeled) human TNF. Subsequently, the wells were treated with different concentrations of Triton X-100, methylene blue, methylene green, basic blue 24 and brilliant cresyl blue and after incubation for 15 minutes, 4 h or 18 h, washed with PBS/BSA. The radioactivity remaining in the individual wells was counted in a Gamma 5500B counter. The results are summarized in figure 9.

The destabilising effect of methylene blue, methylene green and brilliant cresyl blue on labeled human TNF was independently confirmed by gel filtration (figure 10), proving that the assay is indeed a valuable tool for screening stabilising or destabilising compounds:

Radiolabeled huTNF (200 ng/ml) was incubated in PBS/BSA or in PBS/BSA containing one of the above-mentioned agents at 100 μg/ml, for 1 hour at 22°C. After size exclusion chromatography on a Sephacryl S-100 column, the radioactivity present in the different fractions (400 μl) was measured using a gamma-counter. Radiolabeled murine TNF (80 ng/ml) was incubated for 16 hours with biotinylated murine TNF (250 ng/ml) and subsequently allowed to bind on wells coated with extravidin, a modified avidin reagent (Sigma nr E2511) (1 μg/ml). After washing the wells with PBS/BSA, the wells were incubated with PBS/BSA, methylene blue, or with a monoclonal antibody against murine TNF, i.e. 1 F3F3 or TN3. After incubation for 15 minutes, 4 h or 18 h, the wells were washed with PBS/BSA and the radioactivity remaining in the individual wells was counted in a Gamma 5500B counter (figure 11).

The destabilising effect of methylene blue (265 μM) on murine TNF was assessed by gel filtration of radiolabeled murine TNF (120 ng/ml [=2.4 nM]) (figure 12).

Table 1. Effect of muTNF75 mutein on the biological activity of muTNF.

MuTNF was serially diluted in DMEM/FCS, with or without muTNF75¹, and directly or after a preincubation of 24 h at 37°C, 100 μl was added to wells containing L929 cells². After 18 h incubation in the presence of actinomycin D, surviving cells were measured by MTT staining. The biological activity of the samples is represented in U/ml. One U/ml is the concentration of TNF at which half of the cells are killed. Results are means (+/- SD) of three experiments.

Footnotes: 1 : Mutein muTNF75 interacts selectively with TNF-R2. 2: The cytotoxic activity on L929s cells is mediated by TNF-R1.

REFERENCES

Aderka, D., Engelmann, H., Maor, Y., Brakebusch, C. And Wallach, D. 1992. Stabilization of the bioactivity of tumor necrosis factor by its soluble receptors. J Exp Med, 175, 323 - 329.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhan, J., Zhan, Z., Miller, W and Lipman, J. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res, 25, 3389 - 3402.

Banner, D.W., D'Acry, A., Janes, W., Gentz, R., Schoenfeld, H.J., Broger, C, Loetscher, H. and Lesslauer, W. 1993. Crystal structure of the soluble human 55 kd TNF receptor-human TNF beta complex: implications for TNF receptor activation. Cell, 73, 431 - 445.

Camussi, G. and Lupia, E. 1998. The future role of anti-tumour necrosis factor (TNF) products in the treatment of rheumatoid arthritis. Drugs, 55, 613 - 620. - Corti, A., Fassina, G., Marcucci, F., Barbanti, E. and Cassani, G. 1992. Oligomeric tumour necrosis factor alpha slowly converts into inactive forms at bioactive levels. Biochem J, 284 (Pt3), 905 - 910.

Hlodan, R. and Pain, R.H. 1995. The folding and assembly pathway of tumour necrosis factor TNF alpha, a globular trimeric protein. EurJ Biochem, 231 , 381 - 387. - Lucas, R., Heirwegh, K., Neirynck, A., Remels, L., Van Heuverswyn, H., and De Baetselier, P. 1990. Generation and characterization of a neutralizing rat anti-rmTNF- alpha monoclonal antibody. Immunology 71, 218-23.

Mosmann, T. 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Journal Of Immunological Methods 65, 55-63.

Sandbom, W.J. and Hanauer, S.B. 1999. Antitumor necrosis factor therapy for inflammatory bowel disease: a review of agents, pharmacology, clinical results, and safety. Inflamm Bowel Dis, 5, 119 - 133.

Sheehan, K. C, Ruddle, N. H., and Schreiber, R. D. 1989. Generation and characterization of hamster monoclonal antibodies that neutralize murine tumor necrosis factors. Journal Of Immunology 142, 3884-93.

Smith, C.A., Farrah, T. and Goodwin, R.G. 1994. The TNF receptor superfamily of cellular and viral proteins: activation, costimulation, and death. Cell, 76, 959 - 962. Stevens, H. P., van der Kwast, T. H., van der Meide, P. H., Vuzevski, V. D., Buurman, W. A., and Jonker, M. 1990. Synergistic immunosuppressive effects of monoclonal antibodies specific for interferon-gamma and tumor necrosis factor alpha. A skin transplantation study in the rhesus monkey. Transplantation 50, 856-61. - Tartaglia, L. A., Weber, R. F., Figari, I. S., Reynolds, C, Palladino, M. A. J., and Goeddel, D. V. 1991. The two different receptors for tumor necrosis factor mediate distinct cellular responses. Proc Natl Acad Sci 88, 9292 - 9296. - Tribouley, C, Wallroth, M., Chan, V., Paliard, X., Fang, E., Lamson, G., Pot, D., Escobedo, J. and Wiliams, L.T. 1999. Characterization of a new member of the TNF family expressed on antigen presenting cells. Biol Chem 380, 1443 - 1447.

Van Ostade, X., Vandenabeele, P., Everaerdt, B., Loetscher, H., Gentz, R., Brockhaus, M. Lesslauer, W., Tavernier, J., Brouckaert, P., and Fiers, W. 1993. Human TNF mutants with selective activity on the p55 receptor. Nature 361(6409), 266-9

Claims

Claims

1. An artificial, biologically active hetero-oligomeric protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit can form a biologically active homo-oligomeric protein complex on its own, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero-oligomeric protein complex, and whereby said unmodified subunit as well as said modifications show a distinctly different biological activity as monomeric molecule, compared to the oligomeric complexes.

2. The artificial, biologically active hetero-oligomeric protein complex of claim 1 , whereby said unmodified subunit as well as said modifications are biologically inactive as monomeric molecule.

3. The artificial, biologically active hetero-oligomeric protein complex according to the claims 1 or 2, whereby said hetero-oligomeric and said homo-oligomeric protein complex are composed of the same number of subunits.

4. The artificial, biologically active hetero-oligomeric protein complex according to claim 3, whereby hetero-oligomeric protein complex is a heterotrimeric protein complex.

5. The artificial, biologically active hetero-oligomeric protein complex according to claim 4, whereby said subunit is a subunit of a member of the TNF superfamily.

6. The artificial, biologically active hetero-oligomeric protein complex according to claim 5, whereby said subunit is a human or murineTNF subunit.

7. Use of a biologically active hetero-oligomeric protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit forms a biologically active homo-oligomeric protein complex, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero- oligomeric protein complex, and whereby said unmodified subunit as well as said modifications show a distinctly different biological activity as monomeric molecule, compared to the oligomeric molecules, to screen compounds that stabilise or destabilise said hetero-oligomeric protein complex.

8. The use of a biologically active hetero-oligomeric protein complex according to claim 7, whereby said hetero-oligomeric protein complex is an artificial hetero- oligomeric protein complex.

9. The use of a biologically active hetero-oligomeric protein complex according to claim 8, whereby said hetero-oligomeric complex comprises a human TNF subunit as unmodified subunit, and a murine TNF subunit as modification, or a murine TNF subunit as unmodified subunit, and a human TNF subunit as modification..

10. Method to screen compounds that stabilize or destabilize a biologically active hetero-oligomeric protein complex, said protein complex comprising one or more modifications of one subunit, possibly in combination with one or more unmodified forms of the same subunit, whereby said unmodified subunit can form a biologically active homo-oligomeric protein complex on its own, whereby the biological activity of said homo-oligomeric protein complex is essentially the same as that of said hetero- oligomeric protein complex, and whereby said unmodified subunit as well as said modifications have a clearly distinct biological activity as monomeric molecule, compared to the oligomeric complexes, said method comprising a) binding one or more of the subunits in a specific way; b) contacting said hetero-oligomeric protein complex with one or more of said compounds; c) measuring the amount of one or more subunits present in monomeric and/or oligomeric form.

11. The method according to claim 10, whereby said hetero-oligomeric protein complex is artificial.

12. A method according to claim 11 , whereby said hetero-oligomeric protein complex comprises a human TNF subunit as unmodified subunit, and a murine TNF subunit as modification, or a murine TNF subunit as unmodified subunit, and a human TNF subunit as modification.

13. A compound, isolated with the use of a biologically active hetero-oligomeric protein complex according to any of the claims 7-9.

14. A compound, isolated with the method according to any of theclaims 10-12.

15. A pharmaceutical composition, comprising a compound according to claim 13 or 14.