WO2004031240A1 - Monoclonal antibody anti-c3-2 directed against the third component of complement (c3) and its use in methods to inhibit complement activation - Google Patents

Abstract

The invention describes a method to inhibit inflammatory reactions in vivo, more specifically the activation of the complement system. The invention consists of the identification and inhibition of a novel functional domain on the native third component of complement, C3, which domain is essential for the activation of C3. Inhibition of conformational changes of the identified domain prevents the activation of C3, and hence the generation of biologically active peptides such as C3a and C5a, and the formation of membrane attack complexes. The preferred inhibitor is a monoclonal antibody (mAb), a humanised monoclonal antibody or a human monoclonal antibody against the identified domain, or functional fragments derived therefrom, or peptides complementary to the identified domain

Description

MONOCLONAL ANTIBODY ANTI-C3-2 DIRECTED AGAINST THE THIRD COMPONENT OF COMPLEMENT (C3) AND ITS USE IN METHODS TO INHIBIT COMPLEMENT ACTIVATION

Field of the Invention

This invention is in the field of immunology/biochemistry, and describes a method to inhibit inflammatory reactions in vivo, more specifically the activation of the complement system. The invention consists of the identification and inhibition of a novel functional domain on the third component of complement, C3, which domain is essential for the activation of C3. Inhibition of conformational changes of the identified domain prevents the activation of C3, and hence the generation of biologically active peptides such as C3a and C5a, and the formation of membrane attack complexes. The preferred inhibitor is a monoclonal antibody (mAb), a humanised monoclonal antibody or a human monoclonal antibody against the identified domain, or functional fragments derived theref om, or peptides complementary to the identified domain.

Background of the Invention

Activation of the complement system plays a key role in the normal mflammatory response to injury. This system consists of a set of proteins, which circulate in blood as inactive precursor proteins, also known as factors. During activation of the system one factor activates the subsequent one by limited proteolysis and so on. This activation process resembles a cascade system, and, therefore, the complement system is also considered as one of the major plasma cascade systems, the other being the coagulation, the fibrinolytic and the contact systems. The physiological role of the complement system is to defend the body against invading micro-organisms.

The complement system can be activated via three pathways, the classical, the mannose- binding lectin (MBL) and the alternative pathway, all activating a common terminal pathway leading to the formation of the membrane-attack complex (Walport MJ, 2001; Fujita T., 2002; Turner, 1996; Cooper N.R., 1985; Muller-Eberhard H.J. et al, 1980; Muller-Eberhard H.J., 1992, In: Gallin JI, Goldstein IM, Snyderman R (eds): Inflammation: Basic Principles and Clinical Correlates, New York, Raven Press Ltd, p.33). Following complement activation, proinflainmatory peptides like anaphylatoxins C3a and C5a are generated and the membrane attack complex, C5b-9, is formed. Complement activation products, especially the anaphylatoxines, elicit a number of biological effects such as chemotaxis of leukocytes, degranulation of phagocytic cells, mast cells and basophils, smooth muscle contraction and the increase of vascular peπneability (Hugh, 1986). In addition, generation of toxic oxygen radicals and the induction of synthesis and release of arachidonic acid metabolites and cytokines lead to the amplification of the inflammatory response.

Although complement is an important line of defence against pathogenic organisms, its activation can also lead to host cell damage. Complement-mediated tissue injury has been reported in a wide variety of inflammatory diseases, including sepsis and septic shock, toxicity induced by the in vivo administration of cytokines or rnAbs, ήnmune complex diseases as rheumatoid arthritis, systemic lupus erythematosus and vasculitis, multiple trauma, ischaemia- reperfusion injuries, myocardial infarction, and so on. The pathogenic role of complement activation in these conditions is likely related, in some way or another, to the aforementioned biological effects of its activation products. Inhibition of complement activation may therefore be beneficial in these conditions. C3, the third complement component, is known to play a central role in all of the three pathways of complement activation. Human C3 is a 190-kD glycoprotein consisting of an enchain (110 kD) and a β-chain (77 kD), held together by disulpbide and non-covalent bonds (Lambris JD, 1988; Janotova J, 1986; De Brain MHL, et al. 1985). Its normal plasma concentration is 1.0-1.6 g/L (Muller-Eberhard H.J., 1992, In: Gallin JI, Goldstein IM, Snyderman R (eds): Inflammation: Basic Principles and Clinical Correlates, New York, Raven Press, p.33). C3 can be synthesised by many cells, of which hepatocytes are the main producers of plasma C3 (Alper CA, et al. 1969). C3 is synthesised as a single chain precursor, pro-C3, which proteolytically is processed into its two peptide-chains (Morris KM, et al. 1982). The complete amino acid sequence of C3 has been derived from the complete cDNA coding sequence (De Bruin MHL, et al. 1985).

Each of the three complement pathways generates a C3 convertase by a different route, nl. classical pathway C3 convertase (C4b2a), alternative pathway C3 convertase (C3b,Bb) and MBL pathway C3 convertase (C4b2a). Two of the three convertases are identical and all are homologous and have the same activity, nl. proteolytic activation of C3, thereby generating C3a and C3b. As a consequence the principal effector molecules of complement activation and the late events are the same for all three pathways.

Proteolytic activation of C3 leads to cleavage of C3 onto an anaphylatoxic peptide C3a and an opsonic f agment C3b. Covalent attachment of metastable C3b to target cells undergoing complement attack results in generation of C5a and formation of C5b-9 membrane attack complex. The tissue injury that results from complement activation is directly mediated by the membrane attack complex, C5b-9, and indirectly by the generation of anaphylatoxic peptides C3a and C5a. These peptides induce damage through their effect on neutrophils and mast cells. Upon stimulation with C5a, neutrophils produce a serine elastase that causes tissue injury. C5a also triggers the generation of toxic oxygen-derived free radicals from neutrophils, and both C3a and C5a stimulate rapid and enhanced production of leukotrienes from IL-13-primed basophils.

The activation of complement in vivo is tightly regulated at several levels. Both plasma and membrane proteins provide regulation at the levels of C3 and C4 involvement. The plasma protein inhibitors are factor H and C4 binding protein, and the regulatory membrane-bound proteins located on cell surfaces are CR1, DAF and MCP. The proteins with inhibitory activity prevent the release of the anaphylatoxic peptides C3a and C5a by inhibiting the C3 and C5 convertases, by promoting dissociation of the multisubunit complexes, and/or by inactivation of the complexes by proteolysis (Sahu et al., 1998; Campbell et al., 1988).

Binding of most of these proteins to C3 has been shown to be largely determined by the conformation of C3, which conformational state is dependent on the state of the internal thioester. C3 (or C3b) with a cleaved thioester, but not native C3 with an intact thioester, is able to interact significantly with factor B and the other proteins just mentioned (Fishelson Z, 1991; Muller-Eberhard H.J., 1992, In: Gallin JI, Goldstein IM, Snyderman R (eds): Inflammation: Basic Principles and Clinical Correlates, New York, Raven Press, p.33). In agreement herewith are observations that the thioester in C3b is much more labile than that in native C3 (Tack BF, et al. 1980; Law SK, et al. 1981).

Studies with fragments of C3, synthetic C3 peptides, and with mAbs against the various moieties of C3, have enabled the precise identification of a number .of these interaction sites on C3 (Lambris JD, 1988; Fishelson Z, 1991; Becherer D, et al. 1989). However, until now, no monoclonal antibody specifically binding a new functional domain or binding site, which is expressed on the native C3, and less on C3 (C3b) with a disrupted thioester, has been identified.

As has been described earlier, complement-mediated pathology has been reported in several diseases. It is well known that activation of one pathway (classical, alternative or lectin binding pathway) leads to recruitment of the others. For example, activation of the classical pathway results in activation of the alternative pathway. Similarly, activation of the lectin pathway supports the activation of the alternative pathway. Thus, in most clinical conditions multiple pathways are activated. These results suggest the usefulness of a complement inhibitor that blocks all three pathways. The three pathways converge at the C3 activation step. Blocking this step would result in total shutoff of the complement cascade including the generation of C3a, C5a and MAC formation. Attempts have been made to efficiently inhibit complement activation by application of endogenous soluble complement inhibitors (Cl -inhibitor, recombinant soluble complement receptor 1 -rsCRl) or by the aώrinistration of antibodies thereby blocking key proteins of the cascade reaction. Other methods for inhibiting complement can be achieved by neutralising the action of complement derived anaphylatoxin C5a, by interfering with complement receptor 3 (CR3, CD18/l lb)-mediated adhesion of inflammatory cells to the vascular endothelium or by incorporation of membrane-bound complement regulators (DAF-CD55, MCP-CD46, CD59) (Kirschfink, 1997).

Several pharmacologic agents, more particular synthetic peptide inhibitors, that regulate or modulate complement activation have been identified by in vitro assays, but most have been shown to be toxic or of low activity in vivo (Sahu et al., 1998; Glover et al., 1988; Kam et al., 1992; Aoyama et al, 1984).

Most of the anti-C3 mAb identified thusfar, bind on neodeterminants present on activation products of C3 but not on native C3. Most of the binding sites are cryptic in native C3 and become exposed only upon fragmentation, being usually restricted to a certain fragment or fragments (Becherer et al, 1992; Fishelson Z, 1991; Becherer et al., 1989; Hack et al., 1988; WO 8706344). These mAbs are therefore not able to inhibit the proteolytic cleavage of native C3 into C3a and C3b.

Saliu et al., 1996 have described a small molecule inhibitor of C3, called " compstatin" , which was isolated from a phage-display peptide library by screening for binding to C3b. Compstatin also binds to native C3, thereby preventing activation of C3. The present invention however clearly identifies the presence of a new functional domain exposed on native C3, which is essential for the activation and proteolytic cleavage of C3 and is different from the binding site of compstatin. The identification of said new functional domain makes it possible to provide for new and alternative compounds capable of inhibiting complement activation at a stage of the complement cascade where all pathways merge into a central effector pathway, but where most biologically active, complement-derived peptides and macromolecular complexes, such as C3a, C5a and MAC, still have to be generated.

Summary of the Invention

It has now been found that native C3, the third component of complement, contains a novel functional domain, which is essential for the activation and proteolytic cleavage of C3 into C3a and C3b. This domain, in part located on the 23kD-o⁽-chain-fragment of C3c, is expressed preferentially on native C3. Blocking this domain of C3 prevents the generation of

C3a and C3b, and the activation of C5 and subsequent factors.

Therefore, the present invention anus at providing a molecule binding a functional domain expressed on native C3, which is capable of inhibiting the generation of biologically active peptides such as C3a and C3b, thereby inhibiting complement activation. Exemplary molecules may be a mAb against the identified functional domain of C3, or a humanised or human mAb against the identified functional domain of C3, or peptides complementary to the identified functional domain of C3. The present invention also contemplates the use of said molecule for inhibiting complement activation.

The present invention further aims at providing molecules capable of inhibiting the activation of native C3 by binding to the identified new functional domain for use in the treatment of complement-mediated diseases. Furthermore, the present invention aims at providing a prophylactic or therapeutic method to inhibit activation of complement in vivo, which method comprises administering an inhibitor of the novel functional domain of C3.

Another aim of the present invention is to provide for a prophylactic or therapeutic method to treat complement-mediated diseases, which method comprises administering a molecule as described above capable of neutralising the novel functional domain of C3, thereby inhibiting complement activation.

The present invention aims at providing pharmaceutical compositions comprising a molecule capable of inhibiting the generation of biologically active peptides such as C3a and C3b, thereby inhibiting complement activation.

The present invention also aims at providing methods for the preparation of said molecules capable of inhibiting C3 activation.

The invention will be more fully understood after a consideration of the following detailed description of the invention.

Brief Description of the Drawings

Figure 1. Inhibition of the hemolytic activity of complement via the classical [A] or alternative pathway [B] by mAb anti-C3-2 or its F(ab)₂ fragments in human serum. Mab anti-C3-l l , specifically reacting with C3d, is used as control mAb. Cells used in [A] are antibody-sensitized sheep red blood cells and in [B] rabbit erythrocytes. Results are expressed as % lysis of the cells. The indicated mAb anti-C3-2 concentration represents that in the sample added to the sample of fresh serum. Insets show the dose response-curves of serum in the respective assay. The volume of serum in [A] was 1.7 μl and in [B] 23 μl.

Figure 2. Inhibition by mAb anti-C3-2 of C3 fixation to aggregated IgG fixed onto a solid- phase.

Aggregated IgG were fixed onto tubes and incubated with fresh serum. Fixation of C3 to the IgG was measured with ¹²⁵I-labeled anti-C3c antibodies. [A] Inhibition by mAb anti-C3-2 observed with 2 μl of serum; [B] Dilutions of fresh normal human seram (ranging from 5 to 0.3 μl) in the absence of antibodies were tested in the absence or presence of EDTA for control. The results are expressed as % of the amount of antibodies added that had bound to the tubes.

Figure 3. Influence of mAb anti-C3-2 of the generation of C3b/bi/c (A), C4b/bi/c (B) and soluble C5b-C9 (C) complexes in human serum by aggregated IgG (AHG). Results are expressed as nmol/L (A and B) or μg/ml (C). Fresh serum samples incubated with either 2 vol of PBS (buffer control), or 2 vol of PBS and one vol of 0.2 M EDTA (negative control), or one vol of mAb anti-C3-2 and one vol of PBS (antibody control), or one vol of AHG and one vol of PBS (open bar), or one vol of AHG and one vol of mAb anti-C3-2 (closed bar). The reaction was stopped by addition of one vol of 0.2 M EDTA.

Figure 4. Influence of mAb anti-C3-2 of the generation of C3b/bi/c (A), C4b/bi/c (B) and soluble C5b-C9 (C) complexes in human serum by zyrhosan. Results are expressed as nmol/L (A and B) or μg/ml (C). The experiments were performed as described in the legend to figure 3, except that zymosan instead of AHG was used. Figure 5. Influence of mAb anti-C3-2 of the generation of C3b/bi/c (A), C4b/bi/c (B) and soluble C5b-C9 (C) complexes in human seram by E.coli bacteria in human seram. Results are expressed as nmol/L (A and B) or μg/ml (C). Experiments are similar as those described in Figure 3 except that E.Coli, instead of AHG, were tested.

Figure 6. Inhibition by mAb anti-C3-2 of the generation of C3a by E.coli in human serum. Results are expressed as nmol/L. Fresh serum samples incubated with either one vol of mAb anti-C3-2 and one vol of E.coli, or one vol of PBS and one vol of E.coli, or with 2 vol of PBS, were tested for the presence of C3 a by radioimmunoassay after the reaction had been stopped by the addition of one vol of EDTA. A serum sample, which was first incubated with one vol of E.coli for 30 min at 37°C, after which one vol of mAb anti-C3-2 and one vol of EDTA were added, served as control.

Figure 7. Dose-response effect of mAb anti-C3-2 on the generation of C3b/bi/c in human seram by AHG, zymosan, E.coli, or Cobra Venom Factor (CoVF). Results are expressed as nmol C3b/bi/c per L.

Figure 8. Cleavage of the α-chain of human C3 by trypsin is not prevented by mAb anti-C3-2. C3 was digested by limited amounts of trypsin in the presence or absence of mAb anti-C3-2. Lanes 1 and 2: high and low molecular weight markers, respectively; lane 3: C3 incubated with trypsin only (SBTI added after incubation); lanes 4 and 5: C3 incubated with trypsin in the presence of 25 or 250 μg mAb anti-C3-2, respectively; lane 6: C3 incubated with trypsin in the presence of SBTI; lane 7: C3 alone.

Figure 9. Cleavage of the cx-chain of human C3 by the purified C3 -convertase, C3(H₂O)Bb, is prevented by mAb anti-C3-2. Purified C3, in the presence or absence of mAb anti-C3-2, was incubated with factors B and D. Lane 1: high molecular weight markers; lanes 2-5: the mixtures incubated in the absence of mAb anti-C3-2 for 0, 5, 15 and 60 min, respectively; lanes: 6-9: the mixtures incubated in the presence of mAb anti-C3-2 for 0, 5, 15 and 60 min, respectively; lanes 10-13: mAb anti-C3-2, factor B, C3b and C3, respectively. Figure 10. The epitope for mAb anti-C3-2 is 10-fold better expressed on native C3 than on C3 with a disrupted thioester. Binding of ¹²⁵I-anti-C3-2 to C3-Sepharose was inhibited by fluid phase C3 species. Results are given as % of the added ¹²⁵I-labeled anti-C3c antibodies that were bound.

Figure 11. MAb anti-C3-2 does not prevent the conformational changes of C3 following disruption of the thioester. Plasma C3 incubated with methylamine in the presence of mAb anti- C3-2 was tested in the radioimmunoassay for iC3. Results are expressed as % of the input of labeled antibodies bound to the Sepharose in the assay.

Figure 12. The epitope for mAb anti-C3-2 is in part located on the 23kD-α-chain fragment of C3c. [A] Shows an immunoblot of C3 species (and of C4 as a control) incubated with mAb anti-C3-2. Electrophoresis was done under reducing conditions. [B] Peptide-chain specific ELISA. Chains of C3c were purified by preparative SDS-PAGE, fixed onto ELISA plates and incubated with mAb anti-C3-2.

Figure 13. Inhibition of activation of C3 by pre-treatment with mAb anti-C3-2 in 2 baboons challenged with a lethal dose of E.coli. Levels of C3b/bi/c were assessed in blood samples collected at T+0 (i.e. the start of the E.coli infusion), +30, +60, +120, +180, + 240, + 300, + 360 and + 1440 minutes. Open symbols represent levels in the anti-C3-2 treated baboons, filled symbols are mean and SD of 6 baboons that did not receive mAb anti-C3-2.

Figure 14: Nucleic acid and amino acid sequence of the variable kappa light region of the anti-C3-2 mAb (A) and the variable heavy region (B) of the anti-C3-2 mAb.

Figure 15: Binding of different concentrations of C3 (present in recalcified plasma) to the anti-C3-2 antibody immobilised onto the sensorchip. Association and dissociation rate constants were calculated for all separate binding curves. Mean values are: k_a ^■= 9.0 x 10⁴ MV; k_d = 3.5 x W⁴ s^"1 ; K_D = 3.9 x 10^"9 M. Figure 16: Interaction of C3-2 with methylamine-treated C3 (in crude plasma).: A: interaction of C3-2 with 0.4M methylamine-treated plasma; B. interaction of C3-2 with 1.2M methylamine-treated plasma; C. interaction of C3-2 with active C3 in plasma; D. control sample: running buffer (hepes buffered saline) containing 0.4M methylamine.

Figure 17: Amino acid sequence of the mouse anti-C3 heavy and light chain variable domains: Italic: primer-induced sequence; : Sequence confirmed by N-terminal protein sequence analysis; : Sequence confirmed by 'nested' -PCR; VL/VH: Consensus sequence; CDR prediction.

Figure 18: Augments of mouse and humanised anti-C3 heavy and light chain variable domains: Mouse: CDR prediction; Humanised: Humanised resiudes.

Figure 19: Interaction of humanised anti-C3-2 mAb with native C3. The figure shows the binding of different concentrations of C3 (present in recalcified plasma) to the humanized C3-2 antibody immobilized onto the sensorchip. Association and dissociation rate constants were calculated for all separate binding curves. Mean values are: k_a = 3.9 x 10⁴ M-¹ s-'; k_d = 2.3 x 10^"4 s^"1; K_D = 5.9 x 10^"9 M.

Figure 20: Anti-C3-2 antibody prevents human complement-mediated damage of the rabbit isolated heart.

Detailed Description of the Invention

Several patents/patent applications and scientific articles are referred to below that discuss various aspects of the materials and methods used to realise the invention. It is intended that all of the references be entirely incorporated by reference.

Basic for the present invention is the realisation that the third component of complement, C3, contains a novel functional domain which is essential for the activation of C3 and that of the subsequent complement cascade, and that activation of C3 can be inhibited by binding of a mAb to this novel functional domain. However, by no means should this invention be constructed so narrowly, virtually every method to inhibit the function of the identified novel domain on C3 is intended to come into the scope of this invention.

According to the main embodiment, the present invention relates to an inhibitor of complement activation characterised in that said inhibitor specifically binds on a functional domain expressed on native human C3, thereby inhibiting the generation of biologically active peptides such as C3a and C3b, inhibiting the activation of C5 and subsequent factors. The new functional domain is located on the 23kD-^α-chain fragment of C3c, as indicated in the examples, and is clearly different from the binding site of compstatin (data not shown).

The term "inhibitor" as used throughout the invention refers to "a molecule capable of inhibiting" . Both terms are used interchangeably. As used herein, the term "molecule" encompasses, but is not limited to, an antibody and fragments thereof, a diabody, a triabody, a tetravalent or other multivalent antibody, a peptide, a low molecular weight non-peptide molecule (also referred to as "small molecules") specifically binding said functional domain on native C3.

The term "a molecule specifically binding on the functional domain" refers to a molecule, which is capable of forming a complex with the functional domain in an environment where other substances in the same environment are not complexed with the same functional domain.

In a preferred embodiment of the invention the inhibitor is an antibody binding the functional domain expressed on native C3, which is essential for the activation of C3. The term "antibody" refers to polyclonal antibodies, monoclonal antibodies, antibodies

^' which are derived from a phage library, humanised antibodies, human antibodies, syntlietic antibodies, chimeric antibodies, antibody fragments such as, but not limited to single-chain Fv's, or constructs thereof. The term "monoclonal antibody" refers to an antibody composition having a homogeneous antibody population. The term is not intended to be limited by the manner in which it is made. A monoclonal antibody typically displays a single binding affinity for a particular polypeptide with which it immunoreacts. Preferably, the monoclonal antibody used is further characterised as immunoreacting with a specific polypeptide.

More particular said antibody is the murine monoclonal antibody anti-C3-2 produced by a hybridoma that is deposited according to the Budapest Treaty at the collection of DSMZ under the accession number DSM ACC2562. However, it will be apparent to those skilled in the art that the antibodies of the present invention, such as anti-C3-2, may be altered as described further, without affecting their functional activity, i.e. neutralising the functional domain of native C3, and thereby inhibiting C3 activation.

The antibodies of the present invention can be prepared by using a technique which provides for the production of antibody molecules by continuous cell lines in culture. These include but are not limited to the hybridoma technique originally described by Kohler and Milstein (Kohler and Milstein, 1975).

The preparation of high titer neutralising polyclonal antibody can be achieved by immunising rabbits, sheeps, goats, horses or other species, and employing one of several ii munization schemes including those using peptides or peptide conjugates. As an example, rabbits can be immunised with 25 μg of purified human C3 in complete Freund's adjuvant by subcutaneous injection. The animals are then repeatedly boosted at 3 week-intervals, each booster consisting of a subcutaneous injection of 25 μg of human C3 together with incomplete Freund's adjuvant. About one week following each boost, plasma can be obtained from the animals by plasmaphoresis to yield about 100 ml of plasma each week. After recalcification with 10 mM CaCl₂ and formation of a fibrin clot, seram can be obtained by centrifugation. Polyclonal antibodies can be purified from this antiserum by affinity chromatography using C3 coupled to CNBr-activated Sepharose, according to the procedure described for anti-Cl-esterase inhibitor antibodies (Hack CE, et al. 1981).

Polyclonal anti-C3 antibodies of the present invention can also be generated in vivo by administering C3 protein or fragments thereof in combination with at least one foreign T- cell epitope, as an immunogen. Several methods for inducing an antibody response in vivo against self-proteins have been described in the art. One example is the method described in the international patent application WO 9505849.

Monoclonal antibodies of the present invention can be obtained by isolating immune cells from an animal immunised with human C3, and immortalisation of these cells to yield antibody secreting cell lines such as hybridomas. Cell lines that produce the desired antibodies can be identified by screening culture supernatants for the presence of antibody activity, and by establishment of the effect of the selected antibody on the functional activity of the complement system, and in particular that of C3.

Human C3 isolated according to a variety of purification methods may be used to immunise an appropriate host animal. The preferred purification scheme is that described by

Tack BF, et al. 1976.

A variety of immunization protocols may be employed, and may consist of intravenous, subcutaneous, or intraperitoneal immunization, followed by one or more boosts. A suitable adjuvant is Freund's adjuvant. The precise schedule of administration of the human C3 to the host animal in general is not well defined. The choice of the immunization procedure is more dependent on host animal antibody responses to the administered C3, as measured by a suitable assay (vide infra). A preferred immunization procedure, however, is hyperimmunization with human C3 as described (Hack CE, et al. 1988).

Alternatively, lymphocytes, human or murine or other, may be immunised in vitro, as for example can be achieved via a procedure outlined by Voss B, 1986 and in EPA

8610791.6. Other procedures have also been described: Luben R, et al. 1980; Reading C,

1986; Reading C, 1982.

An alternative approach for immunization comprises the use of synthetic peptides that mimic the domain of C3 essential for activation as recognised by mAb anti-C3-2 (vide infra). The methods of making antibodies against peptides are well-known in the art and generally require coupling of the peptides to a suitable carrier molecule, for example bovine serum albumin or keyhole limpet hemocyanin. The peptides can be made according to procedures well known in the art. The procedure also may use commercially available peptide synthesiser machines. Also the hybrid cell line that produces the antibody may be used as a source of DNA or mRNA encoding the desired antibody, which may be isolated and transferred to cells by known genetic techniques to produce genetically engineered antibody.

Monoclonal antibodies can also be produced in various other ways with techniques well understood by those having ordinary skill in the art. Details of these techniques are described in Antibodies: A Laboratory Manual, Harlow et al. Cold Spring Harbor Publications, p. 726 (1988), or are described by Campbell, A.M. ("Monoclonal Antibody Technology Techniques in Biochemistry and Molecular Biology," Elsevier Science Publishers, Amsterdam, The Netherlands (1984)) or by St. Groth et al. (1980)). These other techniques include, but are not limited to techniques for recombinant production of monoclonal antibodies. Monoclonal antibodies of any mammalian species, including humans, can be used in this invention. Accordingly, the antibodies according to this embodiment may be human monoclonal antibodies. Such human monoclonal antibodies may be prepared, for instance, by the generation of hybridomas, derived from immunised transgenic animals, containing large sections of the human immunoglobulin (Ig) gene loci in the germline, integrated by the yeast artificial chromosomal (YAC) technology (Mendez et al., 1997).

As used herein, the teπn "humanised antibody" means that at least a portion of the framework regions of an immunoglobulin or engineered antibody construct is derived from human immunoglobulin sequences. It should be clear that any method to humanise antibodies or antibody constructs, as for example by variable domain resurfacing (Roguska et al., 1994) or CDR grafting or reshaping (Hurle and Gross, 1994), can be used.

As used herein, the term "chimeric antibody" refers to an engineered antibody construct comprising of variable domains of one species (such as mouse, rat, goat, sheep, cow, lama or camel variable domains), which may be humanised or not, and constant domains of another species (such as non-human primate or human constant domains) (for review see Hurle and Gross (1994)). It should be clear that any method known in the art to develop chimeric antibodies or antibody constructs can be used.

As used herein, the term "fragment" or "fragments" refers to F(ab), F(ab)'2, Fv, scFv and other fragments which retain the antigen binding function and specificity of the parent antibody. The methods for producing said fragments are well known to a person skilled in the art and can be found, for example, in Antibody Engineering, Oxford University Press, Oxford (1995) (1996) and Methods in Molecular Biology, Humana Press, New Jersey (1995). As used herein, the term "single chain Fv", also termed scFv, refers to engineered antibodies prepared by isolating the binding domains (both heavy and light chains) of a binding antibody, and supplying a linking moiety which permits preservation of the binding function. This forms,, in essence, a radically abbreviated antibody, having only that part of the hyper-variable domain necessary for binding the antigen. Deteπnination and construction of single chain antibodies are described in U.S. Patent No. 4,946,778 to Ladner et al.

Additional information concerning the generation, design and expression of recombinant antibodies can be found in Mayforth, Designing Antibodies, Academic Press, San Diego (1993). Regardless the nature of the antibody, polyclonal, monoclonal, or recombinant, it may be purified by standard techniques well known in the art. Most of these techniques use affinity chromatography, often in combination with a precipitation step.

Cell lines that secrete antibody against human C3 can be identified by assaying culture supernatants, ascitic fluid, etc., for the presence of antibody. The preferred screening procedure comprises two sequential steps, the first being identification of hybridomas that secrete mAb against human C3, the second being determination of the ability of the mAb to inhibit activation of C3.

The initial screening step of culture supernatants of hybridomas obtained by fusion of lymphocytes of mice immunised with C3, parts thereof, or with C3 peptides, with an appropriate fusion partner, is preferably done by an ELISA or a RIA. Both assays are known to those skilled in the art, and consist of coupling of human C3 to a solid-phase matrix, and assaying for antibody binding to C3 by a second, labelled antibody. In case peptides are used for immunization, peptides coupled to a solid-phase matrix, also can be used in these assays.

The preferred assay is an ELISA in which purified human C3 (Tack BF, et al. 1976) is used for coating, and which is further carried out according to the procedure described by Smeenk RTJ, et al. (1987). Alternative assays may be those described by Langone J, et al. (1983).

Subsequently, an alternative screening procedure may be used to assess whether the selected antibody may bind C3 in solution. This is achieved by a method in which an anti- immunoglobulin agent is coupled to a solid-phase matrix, and bound antibodies against C3 are specifically detected using labelled purified C3. The preferred RIA procedure for screening of C3 antibodies may be that described by Hack CE, et al. (1988). Alternatively, solutions containing C3 may be incubated with the antibody coupled to a solid-phase matrix via an anti-im unoglobulin reagent. The matrix is then washed, and bound C3 is dissociated from it. The eluted C3 may then be measured by SDS-PAGE followed by Western blotting. In addition, any construct of an antibody or a fragment is also a subject of current invention. As used herein, the term "construct" relates to diabodies, triabodies, tetravalent antibodies, pepta- or hexabodies, and the like, that are derived from an anti-human C3 antibody according to the present invention. Said multivalent antibodies, comprising at least one hypervariable domain from an anti-C3 antibody according to the present invention, can be mono-, bi- or multispecific.

The present invention further also relates to C3 -binding peptides and low molecular weight nonpeptides capable of binding the functional domain on C3 homologous to the epitope of the anti-C3-2 mAb.

Said diabodies, triabodies, tetravalent antibodies, C3-binding peptides and low molecular weight nonpeptide molecules can be produced by the following methods:

1) chemical linkage of anti-C3 antibodies of the present invention or univalent fragments thereof following a method as described by Fanger et al. (1992).

2) genetically engineering of non-covalently-hnked diabodies as described by Holliger et al. (1993) and tetravalent antibodies as described by Pack et al. (1995). 3) genetically engineering of covalently-linked chelating recombinant antibodies as described by Kranz et al. (1995), single chain antibodies fused to protein A or Streptavidin as described by Ito and Kurosawa (1993) and Kipriyanov et al. (1996) and bispecific tetravalent antibodies as described in EP 0 517 024 to Bosslet and Seeman, and Coloma and Morrison (1997). 4) genetically engineering of triabodies as described by Kortt et al (1997).

5) phage display of Ab combinatorial libraries resulting in the production of high-affinity antibodies and screening of random DNA sequence phage display libraries for small antigen-binding peptides as described in US patent numbers 5,403,484 and 5,571,698 and 5,223,409 to Ladner et al, Schultz (1996), Parsons et al. (1996), McGuinness et al. (1996), Hoogenboom (1997) and Georgiou et al. (1997).

6)- generation of hybridomas, derived from immunised transgenic mice, containing large sections of the human immunoglobulin (Ig) gene loci in the germ line, integrated by the yeast artificial chromosomal (YAC) technology, resulting in effective blocking antibodies as described by Mendez et al (1997).

7) rational drug design resulting in the production of low-molecular weight nonpeptide molecules as described by Wiley and Rich (1993), Wendolowski et al. (1993) and

Lybrand (1995).

8) "High Throughput Screening" (HTS) of chemical or natural libraries, resulting in the production of peptides or non-peptides as described by Sarabbi et al. (1996).

As used herein, the term "C3-binding peptides or fragments" refers to any peptide (i.e. a polymer composed of at least two amino acids) which cross-links, or reacts with the functional domain on native human C3 homologous to the epitope of the mAb anti-C3-2. The term "low-molecular weight nonpeptide molecules" refers to any molecule which is not a peptide and which cross-links, or reacts with the functional domain on native human C3 homologous to the epitope of the mAb anti-C3-2. According to another embodiment the inhibitor of the functional domain on native human C3 is any molecule capable of binding a functional domain on native human C3 essential in the activation of C3 which comprises an epitope for the mAb anti-C3-2 or said inhibitor can be any molecule competing with the mAb anti-C3-2 for the binding on said functional domain expressed on native human C3. The small molecule inhibitor of C3, compstatin, is not competing with the mAb anti-C3-2 for binding on native C3, indicating that both compstatin and anti-C3-2 mAb have different binding sites on native C3 (data not shown).

The phrase "molecule competing with the mAb anti-C3-2" means that said molecule has the same or comparable specificity for the functional domain exposed on native

C3 as the mAb anti-C3-2. Methods for determining the specificity of a molecule by competitive inhibition, i.e. solid phase ELISA, can be found in Harlow et al. (1988), Colligan et al. (1992, 1993), Ausubel et al. (1987, 1992, 1993), and Muller R. (1993).

A further embodiment of the present invention relates to an inliibitor comprising the variable region or the humanised variable region of the mAb anti-C3-2 or fragments thereof, capable of binding the functional domain exposed on native C3 thereby inhibiting the activation of C3.

The inhibitors described in the present invention are characterised by their ability to inhibit activation of C3. Said inhibitors can be selected by the assessment of their effect on the hemolytic activity of the complement system in human serum, as well as on the generation of complement activation products by complement activators in serum.

The functional properties of selected C3 inhibitors, more particular C3 antibodies may be tested by adding these to fresh human serum, followed by measurement of the hemolytic activity of the mixture in hemolytic assays. These hemolytic assays are well known in the art. To assess activation of complement via the classical pathway, serial dilutions of fresh human seram are added to a constant number of erythrocytes optimally sensitised with IgG/M antibodies, in the presence of veronal buffered saline (VBS) containing CaCl₂ and Mg Cl₂, and incubated under shaking at 37°C. Thereafter, the intact erythrocytes are pelleted by centrifugation, and hemolysis is assessed by measuring hemoglobin content of the supernatant spectrophotometrically. To assess activation of the alternative pathway, a similar type of assay may be used except that non-sensitised rabbit erythrocytes instead of antibody- sensitised red blood cells, and VBS containing MgEGTA rather than CaCl₂ and MgCl₂, are used.

The effect of the selected C3 inhibitor, more particular C3 antibody, on the generation of complement activation products in human serum can be analysed by adding purified inhibitor, more particular antibody, to serum, followed by incubation at 37°C of the mixture with complement activators such as aggregated IgG, cobra venom factor, E.coli bacteria or zymosan. After this incubation EDTA is added to prevent further activation and the mixture is tested for the presence of complement activation products such as C3a, C4a, C5a, C3b/bi/c, C4b/bi/c or C5b-C9. Assays for these complement activation products are well known in the art and can be obtained commercially. The preferred assays are those described by Hack CE, et al. (1988), Hack CE, et al. (1990) and Wolbink GJ, et al. (1993). The inhibitors described in the present invention are according to another embodiment further characterised by specifically binding on a new functional domain expressed on native human C3.

As indicated in the examples, the new functional domain present on native human C3 is characterised by comprising the epitope of the mAb anti-C3-2, being clearly different from the C3 -convertase binding site and is in part located on the 23 kD-^α-chain fragment of C3c. Neutralising said functional domain results in the inactivation of C3, thereby inhibiting the generation of biologically active peptides such as C3a and C3b, inhibiting the activation of C5 and subsequent factors. Accordingly the present invention also relates to a functional domain exposed on native human C3 characterised in that said functional domain is neutralised by binding an inhibitor of the present invention, thereby preventing the generation of the biologically active peptides C3a and C3b.

Comparison of the amino acid sequence data of human C3, C4, C5 and ^α2- macroglobulin, has revealed that these proteins are homologous, although C5 does not contain the internal thioester sequence (De Bruin et al., 1985; Haviland et al., 1991; Lundwall et al., 1985; Belt et al., 1984; Sottrup- Jensen et al., 1985). Therefore, it is reasonable to expect that domains found to be involved in the activation of C3, may exist on C4 and C5 as well. Strategies to inhibit an activation domain of C3, could therefore, be applicable to C4 and C5 as well.

According to another embodiment the inhibitors of the present invention can be used for the preparation of a medicament for inhibiting C3 activation, thereby inhibiting the generation of the biologically active peptides C3a and C3b, inhibiting the activation of C5 and subsequent factors. As a result the inhibitors of the present invention can be used for the preparation of a medicament for inhibiting complement activation in vivo. The inhibitors can be used alone or in combination with other drags.

In a further embodiment the inhibitors of the present invention can be used alone or in combination with other drags, for the preparation of a medicament to treat a host organism suffering of a complement mediated disease, or at risk with respect to such a disease.

The term "complement-related diseases" relates to, but is not limited to, autoimmune diseases such as experimental allergic neuritis, type II collagen-induced arthritis, myasthenia gravis, hemolytic anemia, glomeralonephritis, rheumatoid arthritis, systemic lupus erythematosus and immune complex-induced vasculitis, adult respiratory distress syndrome, stroke, xenotransplantation, multiple sclerosis, burn injuries, extracorporeal dialysis and blood oxygenation, inflammatory disorders, including sepsis and septic shock, toxicity induced by the in vivo administration of cytokines or mAbs, multiple trauma, ischaemia-reperfusion injuries, myocardial infarction. Thus, in the present invention, patients suffering from a disease involving complement- mediated damage can be administered an effective amount of an inhibitor as described so that complement activation is inhibited. By "effective amount" it is meant a concentration of inhibitor, which is capable of inhibiting complement activation.

Accordingly, the present invention relates to a method for inhibiting complement activation comprising the step of administering an inhibitor as described by the present invention.

The present invention also relates to a method for preventing or treating diseases mediated by activation of complement comprising the step of administering an inhibitor as described by the present invention.

Treatment (prophylactic or therapeutic) will generally consist of administering the inhibitors parenterally, preferably intravenously. The dose and administration regimen will depend on the extent of inhibition of complement activation aimed at. Typically, the amount of antibody given will be in the range of 5 to 20 mg per kg of body weight. For parenteral administration, the inhibitor will be formulated in an injectable form combined with a pharmaceutically acceptable parenteral vehicle. Such vehicles are well-known in the art and examples include saline, dextrose solution, Ringer's solution and solutions containing small amounts of human serum albumin. Typically, the inhibitor will be formulated in such vehicles at a concentration of about 100 mg per ml. In the preferred embodiment of this invention the inhibitor is given by intravenous injection. It will, of course, be understood that intended to come within the scope of this invention is virtually every method of administering C3 inhibitors as described by the present invention, to yield sufficiently high levels either in the circulation or locally.

Another embodiment of the present invention relates to a pharmaceutical composition comprising an inhibitor as described above in a pharmaceutical acceptable carrier. The pharmaceutical compositions according to the invention may be foπnulated in accordance with routine procedures for administration by any route, such as oral, topical, parenteral, sublingual, transdermal or by inhalation. The compositions may be in the form of tablets, capsules, powders, granules, lozenges, creams or liquid preparations, such as oral or sterile parenteral solutions or suspensions or in the form of a spray, aerosol or other conventional method for inhalation. The term 'pharmaceutical acceptable carrier' relates to carriers or excipients, which are inherently nontoxic and nontherapeutic. Examples of such excipients are, but are not limited to, saline, Ringer's solution, dextrose solution and Hank's solution. Nonaqueous excipients such as fixed oils and ethyl oleate may also be used. A preferred excipient is 5% dextrose in saline. The excipient may contain minor amounts of additives such as substances that enhance isotonicity and chemical stability, including buffers and preservatives. According to another embodiment the present invention also relates to the use of the mAb anti-C3-2 for identifying molecules capable of binding the new functional domain on native C3. Said molecules can be identified by their capability of competing with anti-C3-2 mAb for binding on the new functional domain on native C3, in a competition assay.

The present invention will now be illustrated by reference to the following examples, which set forth particularly advantageous embodiments. However, it should be noted that these embodiments are illustrative and are not to be constraed as restricting the invention in any way.

Example I: hmnunization with human C3 or peptide immunogens and the production of hybridomas.

The following describes the immunization of mice with purified human C3 with the aim to isolate lymphocytes from the immunised mice and to produce murine hybridomas. It will be further appreciated that the procedure can be employed to produce antibodies against C3 fragments or peptides.

The preferred procedure for immunization is that described by Hack CE, et al. (1988). Briefly, mice were immunised by repeated intraperitoneal injections of 25 μg of purified human C3 given at three- week intervals. The first C3 gift was mixed with complete Freund's adjuvant, the subsequent with incomplete Freund's adjuvant. Four days after the final boost, spleens were removed from the immunised mice and the splenocytes were fused with the murine myeloma cell-line SP2/0-Agl4, according to the procedure first described by Kohler G, et al. (1975), except that feeder cells were replaced by IL-6, formerly called hybridoma growth factor (Aarden La, et al. 1985). Immunised mice were sacrificed and splenocytes teased from the spleens, and washed in serum free Dulbecco's Modified Eagles medium. Similarly, SP2/0-Agl4 myeloma cells were washed, and added to the splenocytes yielding a 5:1 ratio of splenocytes to myeloma cells. The cells were then pelleted, and the supernatant was removed. One ml of a 40 % (v/v) solution of polyethylene glycol 1500 was then added dropwise over a 60 sec period, after wliich the cells were incubated for another 60 sec at 37°C. Nine ml of Dulbecco's Modified Eagles medium was then added with gentle agitation. The cells were pelleted, washed to remove residual polyethylene glycol, and finally plated at a concentration of 10⁵ cells per well in Dulbecco's Modified Eagles medium containing 10% (v/v) fetal calf serum ( lOOμl per well). After 24 hours, 100 μl of a 2x solution of hypoxanthine/azaserine selection medium was added to each well. At day 4 hypoxanthine/azaserine selection medium was replenished, at day 7 it was replaced by Dulbecco's Modified Eagles medium containing 10% (v/v) fetal calf seram. During all incubations monocyte derived or recombinant human IL-6 was present in the culture at concentrations of approximately 10 pg/ml. About 80% of the wells exhibited cell growth at day 10. Example 2: ELISA for the detection of C3 antibodies.

The wells were screened for the presence of antibody to C3 using an enzyme-linked immuno sorbent assay, in which purified human native C3 was used for coating (2 μg/ml in PBS, pH 7.4 (PBS); 100 μl/well). Residual non-specific binding sites were then blocked by a 30 minutes incubation at room temperature with PBS/0.1% (w/v) Tween 20 (PBS-T) containing 0.2 % (w/v) gelatin (PBS-TG). Then, after a wash procedure (5 times with PBS- T), the plates were incubated for 120 min at 37°C with 20 μl of hybridoma supernatant together with 80μl of PBS-TG. Finally, bound murine antibodies were detected by an incubation with peroxidase-conjugated polyclonal goat anti-mouse immunoglobulin antibodies for 120 min at 37°C. Finally, the plates were washed with distilled water (5 times), and developed with 3,5,3 ',5' tetratmethyl benzidine. In this way, mAb anti-C3-2 was identified.

Example 3: Preparation of purified mAb anti-C3-2 and its F(ab)'₂-fragments.

Antibody may be produced in vitro from the hybridoma anti-C3-2 by culturing the cells in 1 litre roller-bottles in Iscove's Modified Dulbecco medium supplemented with 2% (v/v) fetal calf serum, 10 pg/ml IL-6, 50 μM 2-mercaptoethanol, and penicillin and streptomycin. The cells were grown to a density of > 10⁶ cells per ml, and one to two weeks later the supernatants were collected. Solid ammonium sulphate was added to yield 50% saturation (i.e., approximately 2M), and an antibody-enriched fraction was obtained by centrifugation for 30 min at 1,300 g. The precipitate was dissolved in 1.5 M NaCl/0.75 M glycine, pH 8.9, and put onto a protein A-Sepharose column (Pharmacia). The column was washed with PBS, and then mAb anti-C3-2 was eluted off with glycine-HCl, pH 2.5. Fractions were neutralised instantaneously with 2M TRIS, pH 8.0, and those containing protein were pooled and dialysed against PBS.

F(ab)'₂-fragments of mAb anti-C3-2 were prepared by incubating overnight at 37°C 2 vol of anti-C3-2 in PBS with one vol of 0.1 M sodium acetate containing pepsin (Cooper Biomedical; 2500 U/mg) to yield a final ratio of pepsin to anti-C3-2 of 1:50, and a pH 4.1. Thereafter, one vol of 1 M TRIS, pH 8.1, was added, and the preparation was dialysed against PBS. On SDS-PAGE the preparation appeared to contain approximately 10-20% uncleaved anti-C3-2, which was removed by passing the preparation over a protein A-column (Pharmacia). Pepsin-digestion did not affect the binding of anti-C3-2 to purified C3 coated onto plastic plates as was assessed with an enzyme-linked sorbent assay, in wliich peroxidase- conjugated rat mAb against mouse kappa-light chain was used to detect bound anti-C3-2 mAb or fragments thereof. Using a similar assay with rat mAb against the heavy chain of mouse IgGl, it was demonstrated that the digested preparation contained less then 1% of uncleaved anti-C3-2 mAb.

Example 4: Inhibition of the hemolytic activity of seram and of purified C3 by mAb anti-C3-

2.

The effect of mAb anti-C3-2 on the hemolytic activity of fresh serum was studied in two types of experiments. A) Inhibition of hemolysis initiated via activation of the classical pathway was studied using erythrocytes optimally sensitised with IgG/M antibodies as a lytic target for fresh human seram. One vol of PBS containing varying concentrations of mAb anti-C3-2, was added to one vol of fresh human serum. Then, 34 μl of 1 to 10 dilutions of the mixtures (i.e. 1.7 μl of serum, respectively) were added to VBS/1 mM CaC12/0.15 mM MgC12 (VBS) yielding a final volume of 100 μl. Hundred μl of sheep red blood cells optimally sensitised with rabbit antibodies (3x10^s cells/ml in VBS) were then added and the mixtures were incubated in a shaking water-bath for 60 min at 37°C. After centrifugation of the mixtures, hemolysis was determined spectrophotometricaliy. As is shown in Fig 1 A, mAb anti-C3-2 inhibited this activation in a dose-dependent fashion, yielding 90% inhibition at a concentration of 7.5 μM. This inhibition of complement mediated lysis of the erythrocytes was not due to consumption of complement by aggregated material in the anti-C3-2 preparation, since F(ab)'2-fragments displayed a similar effect. In contrast, mAb anti-C3-l l, tested as a control, was not able to inhibit lysis of the antibody-sensitised red blood cells by fresh serum. Complete inhibition of hemolytic activity by mAb anti-C3-2 was not observed at any of the concentrations tested. Most likely, this residual hemolytic activity represented the C3-bypass by C4 (Masaki T, et al. 1991).

B) Inhibition of hemolysis initiated via activation of the alternative pathway was studied using rabbit erythrocytes (which directly, i.e., in the absence of sensitising agents, activate the alternative pathway in fresh human seram). Mixtures of fresh human seram and mAb anti-C3-2 were prepared as described under A). Forty-six μl of the mixtures (i.e. 23 μl of serum, respectively) were added to VBS containing 10 mM Mg-EGTA, yielding a final volume of 100 μl. Fifty μl of rabbit erythrocytes (2.5 x 10⁸ cells/ml) were then added and the mixtures were incubated in a shaking water-bath for 60 min at 37°C. After centrifugation of the mixtures, hemolysis was determined spectrophotometricaliy. As can be seen in Fig IB, mAb anti-C3-2 inhibited alternative pathway mediated lysis in a dose dependent fashion yielding 90% inhibition of lysis at a concentration of 3.8 μM. In contrast, although intact mAb anti-C3-l l also inhibited this alternative pathway mediated lysis, its F(ab)'₂-fragments did not (Fig IB).

In another set of experiments the effect of mAb anti-C3-2 on the hemolytic activity of purified C3 was investigated. Hundred μl of VBS containing the indicated amount of mAb was added to 300 μl of VBS containing the indicated amount of purified human C3. The mixtures were then incubated for 45 min in an ice-bath. Then, 25 μl of C3 -depleted serum (Hack, CE et al. 1988,), lOμl of purified Clq (100 μg/ml in VBS) and 100 μl sheep erythrocytes (3 x 10⁸ cells per ml of VBS) optimally sensitised with rabbit antibodies, were added, and the mixture was further incubated in a shaking water-bath for 60 min. at 37°C. Then, 3.5 ml of ice-cold 0.15 NaCl was added and the tubes were centrifuged at l,300g for 10 min. Hemolysis was measured spectrophotometricaliy. The results, shown in Table 1, clearly indicated that mAb anti-C3-2 was able to inhibit the hemolytic activity of purified C3 in a dose dependent fashion. For example, nearly 50% lytic activity of C3 was inhibited by pre- incubating 17.7 μg of C3 with 15 μg of mAb anti-C3-2. In contrast, mAbs anti-C3-9 or -24 had no effect on the hemolytic activity of purified C3.

Table 1: monoclonal antibody anti-C3-2 inhibits the hemolytic activity of human C3.

C3 added (μg) 17.7 35 07 mAb added (μg) anti-C3-2

75 16 2 0

15 55 2 0

3 100 7 0 anti-C3-9

75 100 103 96

15 101 105 100

3 99 103 101 anti-C3-24

75 101 102 90

15 99 103 102

3 103 101 96 buffer 101 100 96

Results were corrected for background lysis (15%) and are given as % lysis. (Note that the concentrations ofC3 used in this experiment are sufficient to yield complete lysis of the added sensitised cells.)

Example 5: Inhibition of fixation of C3 to aggregated IgG by mAb anti-C3-2. Aggregated human IgG (AHG; Hack CE, et al. (1981)) was incubated overnight in 2 ml-polystyrene tubes at a concentration of 10 μg/ml in PBS (final volume 0.5 ml). Then, the tubes were washed twice with PBS-T. Thereafter, the tubes were incubated with 500 μl of VBS/0.1% (w/v) Tween 20 (VBS-T) containing 2 μl of fresh normal human serum as well as varying amounts mAb anti-C3-2 or control antibodies, for 2 hours at 37°C. Then, the tubes were washed three times with PBS-T, and radioactivity bound to the tubes was assessed with a gamma-counter. Results were expressed as % binding of input of ¹²⁵ 1-anti-C3c (Fig 2A). As a control, dilutions of fresh normal human seram (ranging from 5 to 0.15 μl) in the absence of mAb were tested (Fig 2B). Binding of ¹²⁵I-anti-C3c in the presence of 10 mM EDTA during the incubation with seram was <2% of the input. It can be seen in Fig 2 A that mAb anti-C3-2, but not control mAb, inhibited the fixation of C3 to aggregated IgG in a dose-dependent fashion.

Example 6: Inhibition by mAb anti-C3-2 of the generation of complement activation products in serum by several complement activators.

A number of experiments were performed to establish whether the generation of C3a, C3b/bi/c and C5b-C9 complexes in serum by complement activators could be inliibited by mAb anti-C3-2. These experiments were all done according to a similar protocol: one vol of mAb anti-C3-2 was mixed with one vol of fresh serum. Thereafter, one vol with complement activator was added and the mixtures were incubated for 20 min at 37°C. Then, one vol of EDTA was added to block further activation, and, finally, the mixtures were tested for the presence of C3b/bi/c, C4b/bi/c, C5b-C9 complexes, and in some experiments also C3a, by assays previously described or obtained commercially (Wolbink GJ, et al. 1993; Hack CE et al. 1988; Behringwerke AG, Marburg, Germany). Fig 3 shows the inhibition by mAb anti- C3-2 of the generation of C3b/bi/c (nM) and soluble C5b-C9 complexes (mg/L) in human serum by aggregated IgG (AHG; at concentrations of 1 and 0.2 mg/ml). One vol of mAb anti- C3-2 (15 μM PBS) was added to one vol of fresh human seram. The mixture was then incubated with one vol of AHG (1 or 0.2 mg/ml in VBS) for 20 min at 37°C. Finally, one vol of 0.2 M EDTA, pH 7.5, was added and serial dilutions of the mixtures were tested for the presence of C3b/bi/c, C4b/bi/c and C5b-C9 complexes. Fresh seram samples incubated with either 2 vol of PBS, or one vol of mAb anti-C3-2 and one vol of PBS, or one vol of AHG and one vol of PBS, as well as fresh serum incubated at 0°C with two vol of PBS and one vol of EDTA 0.2 M, served as controls. The results (Fig 3) showed that mAb anti-C3-2 strongly inhibited the generation of C3b/bi/c and C5b-C9 complexes, but, as expected, not that of C4b/bi/c. Actually, mAb anti-C3-2 did slightly increase levels of the latter activation product in the mixture, presumably do to the presence of some aggregated material in the anti-C3-2 preparation used.

Fig 4 shows a similar type of experiment, using zymosan (Sigma Chem Co., St. Louis, MO) washed with PBS as an activator. A slight modification of this experiment as compared with the previous, was that after the incubation at 37°C and the addition of EDTA, the samples were centrifuged for 10 min at 3,000 g to remove the zymosan particles. Also in this experiment mAb ani-C3-2 inhibited the generation of C3b/bi/c and C5b-C9 complexes in serum by zymosan added at concentrations of 5 or 1 mg/ml.

Fig 5 shows the inhibition by mAb anti-C3-2 of the generation of C3b/bi/c and soluble C5b-C9 complexes in human serum by E.coli bacteria. The experiment was performed in a similar way as those described above, except that E.coli organisms (Hinshaw LB, et al. (1983)) at concentrations of 10¹⁰ or 10⁹ organisms per ml were used to activate the complement system in serum. After the incubation at 37°C and the addition of EDTA, the mixtures were centrifuged for 10 min at 1,300 g to remove the E.coli organisms. Again, mAb anti-C3-2 significantly inhibited the generation of C3b/bi/c and C5b-C9 complexes in seram. Mab anti-C3-2 not only inhibited the generation of C3b/bi/c, but also that of C3a as is shown in Fig 6: one vol of mAb anti-C3-2 (15 μM in PBS) was added to one vol of fresh human serum. Then, one vol of E.coli (10¹⁰ organisms per ml in PBS) was added and the mixture was incubated for 30 min at 37°C. Finally, one vol of 0.2 M EDTA, pH 7.5, was added and serial dilutions of the mixtures were tested for the presence of C3a(desarg) by a specific radioimmunoassay (Hack CE, et al. 1988). Results are expressed in nM. Fresh serum samples incubated with either 2 vol of PBS, or one vol of PBS and one vol of E.coli, as well as a serum sample, which was first incubated with one vol of E.coli for 30 min at 37°C, after which one vol of mAb anti-C3-2 and one vol of EDTA were added, served as controls.

Finally, the dose-response effect of mAb anti-C3-2 on the generation of C3b/bi/c in human seram by aggregated IgG, zymosan, or E.coli was investigated. One vol of varying concentrations of mAb anti-C3-2 (1.9-15 μM in PBS) or one vol of PBS, was added to one vol of fresh human serum. Then, one vol of AHG (0.2 mg/ml), zymosan (1 mg/ml) or E.coli (10⁹ organisms/ml) was added and the mixture was incubated for 20 min at 37°C. One vol of 0.2 M EDTA, pH 7.5, was then added and serial dilutions of the mixtures were tested for the presence of C3b/bi/c by ELISA. Results were expressed as nM C3b/bi/c. As is shown in Fig.7, mAb anti-C3-2 inhibited the generation of this complement activation product in a dose-dependent fashion, independently on the nature of the activator.

In a separate experiment a similar effect of mAb anti-C3-2 on the generation of C3b/bi/c by Cobra Venom Factor in serum was observed (data not shown). Example 7: mAb anti-C3-2 prevents cleavage of the ^α-chain of C3 by a C3 -convertase, but not that by trypsin.

An obvious explanation for the inhibiting effect of anti-C3-2 could be that its epitope (partially) overlapped the cleavage site on the ^α-chain for trypsin, which under limiting conditions also cleaves the bond between arg-748 and ser-749. Therefore, purified human C3 (25 μg; Tack BF, et al. 1976), mAb anti-C3-2 (25 or 250 μg), trypsin (0.25 μg; Sigma Chem Co, St Louis, MO) in PBS (final volume 116 μl) were incubated for 3 min at 37°C. Then, soybean trypsin inhibitor (SBTI; 25 μl containing 25 μg; Sigma), were added. Hundred μl of the mixture was then incubated for 5 min at 100°C with 100 μl of reducing sodium dodecyl sulphate (SDS) sample buffer (Hoekzema R, et al. 1985). Subsequently, 100 μl of these mixtures were electrophoresed on SDS-polyacrylamide gels (5-20 %, w/v). Finally, proteins on the gel were visualised by staining with Coomassie Brillant Blue. The results are shown in Fig 8: Lanes 1 and 2: high and low molecular weight markers, respectively; lane 3: C3 incubated with trypsin only (SBTI added after incubation); lanes 4 and 5: C3 incubated with trypsin in the presence of 25 or 250 μg mAb anti-c3-2, respectively; lane 6: C3 incubated with trypsin in the presence of SBTI; lane 7: C3 alone. The results clearly indicated that mAb anti-C3-2 could not prevent cleavage of the ^α-chain by trypsin (Fig.8; the upper arrow on the left indicates the position of the uncleaved ^α-chain, the lower that of the (cleaved) ^α'-chain). This was not due to the fact that mAb anti-C3-2 was degraded by trypsin, since binding of mAb anti-C3-2 to C3 was not affected by incubation with trypsin under these conditions, as was assessed with an ELISA in which purified C3 was coated on microtiter plates, and peroxidase-conjugated goat-anti-mouse immunoglobulin was used as a detecting antibody (results not shown).

Cleavage of the ^α-chain of human C3 by the purified C3 -convertase, C3(H₂O), Bb, was, however, prevented by mAb anti-C3-2, as is shown in Fig 9: Sixty μg of purified human C3 containing about 6 μg of C3(H₂0), was incubated with 2 μg of purified human factor D and 40 μg of human factor B in the presence or absence of mAb anti-C3-2 (240 μg) in VBS containing 10 mM Mg-EGTA (total volume 500 μl) at 37°C. After 5, 15 and 60 min of incubation, 50 μl samples were taken and mixed with 5 μl 0.5 M EDTA to stop further activation. An equal vol of reducing SDS-sample buffer was then added to the mixtures. After an incubation for 5 min at 100°C, the samples were subjected to SDS-PAGE on 7.5 %, w/v, gels. Proteins were visualised by staining with Coomassie Brillant Blue. The results are shown in Fig 9: Lane 1: high molecular weight markers; lanes 2-5; the mixtures incubated in the absence of mAb anti-C3-2 for 0, 5, 15 and 60 min, respectively; lanes: 6-9: the mixtures incubated in the presence of mAb anti-C3-2 for 0, 5, 15 and 60 min, respectively; lanes 10- 13: mAb anti-C3-2, factor B, C3b and C3, respectively. The conclusion of this experiment was that in the presence of anti-C3-2 there is hardly any cleavage of the ^α-chain of C3 and of factor B. Thus, although mAb anti-C3-2 could not prevent cleavage of the ^α-chain by trypsin, it did prevent cleavage by a C3 -convertase. Therefore, mAb anti-C3-2 does not bind to the C3 -convertase cleavage site, but rather recognises a hitherto unknown domain on C3, which is essential for cleavage by a C3-convertase.

Example 8: Mab anti-C3-2 binds to native C3 and is not able to prevent the conformational changes induced by disruption of the thioester.

A possible explanation for the inhibition of C3 activation by mAb anti-C3-2 could be that this mAb stabilises the native conformation of C3 and prevents the conformational changes occurring after disruption of the thioester. We therefore analysed the expression of the epitope for mAb anti-C3-2 on C3 species with an intact or a disrupted thioester, and also studied whether it could prevent the appearance of neoepitopes following disruption of the thioester using mAb against these neoepitopes (Hack CE, et al. 1988). Mab anti-C3-2 (and as a control mAb anti-C3-l, which is directed against an epitope on the β-chain expressed equally well on activated and non-activated C3 species) was purified from hybridoma supernatant using protein G affinity chromatography following the instructions of manufacturer (Pharmacia), and labelled with ^I25I with the Chloramin T method to yield a specific activity of 5-7 μCi/μg protein. Approximately 2-3 ng of ^I25I-mAb was then incubated with 0.5 ml of a 1 mg/ml suspension of Sepharose 4B to which native C3 was coupled, in the presence of varying concentrations of native C3, C3b (prepared by limited trypsin digestion of native C3), C3c (purified from aged serum as described by Hack CE, et al. 1988, and C3 in which the thioester was disrupted by incubation (60 min at 37°C) with 0.4 M methylamine, pH 7.5 (iC3). After an overnight incubation at room temperature, the beads were washed and bound ¹²⁵I-mAb was assessed by counting the radioactivity bound to the beads using a multichannel gammacounter. The results were expressed as % inhibition (compared to binding of the labelled mAb in the absence of competitor), and are presented in Fig 10. It was observed that mAb anti-C3-l, as expected, bound equally well to the C3 species with a disrupted thioester.

The influence of mAb anti-C3-2 on the conformational changes of C3 following disraption of the thioester, was studied by adding 20 μg of mAb anti-C3-2 in 10 μl PBS (or 10 μl of PBS only, as a control) to 100 μl of fresh human plasma (1 to 50 diluted in PBS or PBS only) and incubating the mixture for 60 min at room temperature. Dilutions of the mixtures were then tested in the radioimmunoassay for iC3 (uncleaved C3 with a disrupted thioester) as described (Hack CE, et al. 1990). Preincubation of plasma with mAb anti-C3-2 did not influence the generation of iC3 in plasma by methylamine (Fig 11). Therefore, we concluded that mAb anti-C3-2 could not prevent the conformational changes following the disraption of the thioester.

This latter observation, as well as the experiment in which it was observed that mAb anti-C3-2 prevents cleavage of the ^α-chain by a C3 -convertase but not that by trypsin (see Example VII), illustrate the unique properties of mAb anti-C3-2, and have never been described for other rnAbs against C3.

Example 9: The epitope for mAb anti-C3-2 is in part located on the 23kD-^αchain fragment ofC3c. Initially we attempted to map the epitope for mAb anti-C3-2 using immunoblotting of

C3 species. Samples of C3 species (5 or 1 μg) purified as described (Hack CE, et al. 1988), were incubated for 5 min at 100°C in reducing SDS sample buffer and thereafter electrophoresed on SDS-polyacrylamide gels (5-20 %, w/v). Purified C4 (Hessing M, et al. 1993) was included as a control. Proteins were then transferred to nitrocellulose sheets. The sheets were incubated in PBS-T containing 5% (w/v) de-lipidated protein powder (Protifar) for 30 min at 37°C, followed by an incubation of biotinylated mAb anti-C3-2 (approximately 2 μg/ml) in the same buffer. The sheets were washed with PBS-T (3x10 min), and subsequently incubated for 60 min at 37°C with streptavidin-peroxidase (Sanquin, Business Unit Reagents, Amsterdam, the Netherlands) at a 1 to 1,000 dilution in the PBS/Tween/Protifar buffer. After a washing procedure the blots were developed with 4- chloro-1-naphtol as described (Westgeest AAA, et al. 1985). The results are shown in Fig 11. Clearly, mAb anti-C3-2 binds to the 23 kD ^α-chain fragment of C3c, corresponding to residues 749 to 955. These results were confirmed by an enzyme linked sorbent assay using peptide-chains purified from C3c as antigens. The 23kD and 45 kD ^α-chain fragments and the β-chain of human C3c were isolated by preparative SDS-PAGE (5-20 %, w/v) of immunopurified human C3c (Hack CE, et al. 1988). Purified fragments, as well as intact C3c for comparison, were then coated onto ELISA plates by incubation overnight at room temperature (100 μl of 5 μg/ml solution in PBS). The plates were then washed with PBS-T, residual binding sites were blocked by incubation with PBS-T containing 3% (w/v) bovine serum albumin for 30 min at room temperature. The plates were then incubated for 60 min at room temperature with 1 μg of biotinylated-mAb (in 100 μl PBS-T), washed (5 times with PBS-T) and incubated with streptavidin-peroxidase, 1 to 1,000 diluted in PBS-T. After another washing (5 times with distilled water), bound antibodies were detected by addition of 3,5,3', 5'-tetramethylbenzidine. The reaction was stopped with 2 M H2SO4. The results, given as absorbance at a wavelength of 450 nm, are shown in Fig 12. It appeared that mAb anti-C3-2 bound to the 23kD fragment of the ^α-chain of C3c, consistent with the results of the immunoblotting. However, also some binding to the β-chain was observed, which was in agreement with the immunoblotting experiments where in some experiments an inconsistent binding to the β-chain was observed (data not shown). Thus, the epitope was at least in part located on the 23 kD fragment of the ^α-chain of C3c.

Example 10: Effect of mAb anti-C3-2 on complement activation in vivo.

Baboons challenged with a lethal dose of E.coli display a pronounced activation of complement in the circulation (De Boer JP, et al. 1993). To asses the ability of mAb anti-C3- 2 to inhibit complement activation in vivo, 2 baboons were pre-treated with mAb anti-C3-2, after which they were challenged with a lethal dose of E.coli. After immobilisation and proper anaesthesia with sodium pentobarbital, two baboons from a mixed breed of Papio c. cynocephalus and Papio c. anubis (Charles River Breeding Laboratories Inc., Wilmington,

USA) were intravenously infused with mAb anti-C3-2 (20 mg per kg body weight) and 30 min thereafter with a lethal dose of E.coli (4x10¹⁰ micro-organisms per kg of body weight) given as a two-hour infusion (Hinshaw LB, et al. 1983). During recovery from shock, the baboons were observed daily and medically treated as appropriate. All 5 control animals died, one of the two anti-C3-2 animals survived. The surviving animal (baboon #2) was euthanized after 7 days with sodium pentobarbital. During the experiments, heart rate, mean systemic arterial pressure, respiration rate and rectal temperature were monitored hourly for six hours and daily for 7 days. Haematologic parameters were assessed in blood samples collected at T+0 (i.e. the start of the E.coli infusion), +30, +60, +120, +180, +240, +360 and +1440 minutes. In addition, at each of these time points also 5 ml blood samples were collected in 10 mM EDTA/100 μg/ml SBTI/10 mM benzamidine (final concentrations). Levels of C3b/bi/c in these samples were determined as described (De Boer JP, et al. 1993). The results, shown in Fig 13, were expressed as % of C3b/bi/c of the standard which consisted of normal baboon serum aged, i.e. normal baboon serum incubated for 7 days at 37°C in the presence of 0.02%, w/v, sodium azide. Compared to the 5 animals that received a similar dose of E.coli but without pre-treatment with mAb anti-C3-2 (given as mean +/-SEM in Fig 13), the pre-treated animals displayed considerably less C3 activation, indicating that mAb anti-C3-2 could attenuate complement activation not only in vitro but also in vivo.

Example 11: Specific interaction of C3-2 with native C3

The C3-2 antibody was shown to interact specifically with native C3, containing an intact thioester bond and not with an inactive form, containing a hydrolysed ester bond. This was shown using Surface Plasmon Resonance-analysis with the BIACORE 3000^® (Biacore AB,

Uppsala, Sweden). This technology permits real-time measurements using surface plasmon resonance (SPR). SPR is an optical phenomenon, seen as a sharp dip in the intensity of light reflected from a metal film coated onto a glass support. The position of the dip depends on the concentration of solutes close to the metal surface. In general, a protein (e.g. antibody) is coupled to the dextran layer (covering the gold film) of a sensor chip and solutions containing different concentrations of a binding protein (e.g. antigen) are allowed to flow across the chip.

Binding (association and dissociation) is monitored with mass sensitive detection.

For determining the interaction between the C3-2 antibody and C3 (native or hydrolysed), BIACORE^® experiments were performed in which the C3-2 antibody was immobilised onto a CM5 sensorchip (Biacore AB). The C3-2 antibody was immobilised using amine coupling according to the manufacturer's procedure. Briefly, the antibodies were diluted to 5 μg/ml in 10 mM acetatebuffer pH4.8 and injected at 5 μl/min until an immobilisationlevel of +/- 500 RU was reached. Injecting 0.1M ethanolamine pH 8.5 blocked residual unreacted ester groups. An irrelevant antibody of the same subclass was immobilised to the same level and was used for subtraction of non-specific binding of C3 to the mlgGl surface.

Native C3 (present in human plasma) was injected at 10-300 μg/ml (concentration active C3 in plasma: 1.2 mg/ml). 150 μl C3 solution was injected at 50 μl/min. The surface was regenerated with 2 pulses of 60 μl 0.2M Na2CO3 pHl l (50 μl/min). Binding to the irrelevant IgGl surface was subtracted as a blank. Kinetic constants were calculated using the BIAevaluation software 3.1. Results of a typical experiment are shown in figure 15. These data show that the C3-2 antibody interacts with high affinity with active C3 present in plasma.

Inactive C3 (iC3) was prepared by treating human plasma with the nucleophilic reagent methylamine. This results in the cleavage of an intramolecular thioester bond and induces C3b-like properties (Pangburn et al., 1981). Treatment was done by incubating 1 volume of plasma with 1 volume of methylamine. HC1 (either 0.4 or 1.2M) during 1 hour at 37°C. The treated plasma was injected over the C3-2 coated sensorchip and binding to the immobilized C3-2 was monitored. Results are shown in figure 16. These data show that treatment with 0.4M methylamine resulted in a decreased association rate combined with no change in dissociation rate, indicating a reduction in concentration of active C3. However, binding was fully abolished upon complete inactivation of C3. This was accomplished when plasma was treated with 1.2M methylamine. These results clearly indicate that C3-2 specifically interacts with C3 containing an intact thioester bond. Example 12: Humanization of anti-C3-2 mAb.

A pellet of approx. 10^δ subcultured hybridoma cells was prepared and total RNA was prepared (QIAGEN Rneasy procedure) for subsequent cDNA synthesis (QIAGEN OneStep RT-PCR) of both variable regions (heavy and light) with 'gene-specific' oligonucleotides. The mouse VH genes were amplified using the primers VH1BACK and VH1-FOR2. These are consensus primers that cover the majority of mouse heavy chain gene families. The primer VK2BACK is used in combination with a mix of four J region primers (MJK1FONX, MJK2FONX, MJK4FONX and MJK5FONX) to amplify the light chain kappa families. The VK2B ACK is a consensus primer that covers most of the mouse kappa families.

PCR fragments of approx. 350 bp were obtained (with VH slightly larger than VL). The specific PCR fragments of the expected size were excised from gel, purified, cloned into pGEM-T vector and sequenced. For the VH, in total 14 clones were sequenced, of which 10 were completely identical. 3 clones differed from the previous 10 at one position within framework regions and 1 clone had an undetermined nucleotide in CDR3. For the VL, in total 16 clones were sequenced, of which 14 were completely identical. 2 clones differed from the previous 14 at several positions or had some undetermined nucleotides.

The first 8 amino acids of the light and heavy chain variable regions were uncertain since the nucleotide sequence at these positions is primer-induced. N-tenrπnal protein sequencing of both antibody chains (up to and including CDR1) demonstrated that a correction was required for residues 1 (D -> E) and 3 (E->V) of the light chain and for residues 1 (* -> D) and 3 (K - > Q) of the heavy chain variable region. Furthermore, the N-terminal protein sequencing confirmed the correct sequence of the residues of the first framework region and CDR1 of both variable regions.

The last 11 and 8 amino acids of resp. heavy and light variable regions also remained uncertain, since the obtained nucleotide sequence in this region is induced by the PCR-primer sequence. Based on the knowledge that the mouse monoclonal C3-2 antibody is an

IgGl/kappa antibody, some extra RT-PCRs with anti-sense primers annealing in the beginning of the constant regions were performed to obtain the correct sequence of the last amino acids of the variable region. PCR fragments of approx. 400 bp were obtained (with VH slightly larger than VL). The specific PCR fragments of the expected size were excised from gel, purified, cloned into pGEM-T vector and sequenced. Sequence analysis on 4 extra clones demonstrated that no correction was required for the last 8 amino acids of the light chain variable region, whereas 1 correction (residue 112: Q -> A) was needed in the last 11 amino acids of the heavy chain variable region.

The finally obtained sequence is shown in figure 17.

We have applied a computer modelling protocol for the construction of humanized C3-2 Fv fragments. The amino acid sequences for the VL and VH domains were humanised using a resurfacing strategy. This method essentially includes the identification and proposal for replacement of residues which (i) significantly differ between the human and mouse consensus sequences, (ii) which are solvent-oriented, and (iii) which are assumed to have no influence on the affinity of the antibody for its antigen. Typical for this approach is that the humanisation consists of 2 main parts. In the first part, a 3D-structure of the mouse Fv is constructed. For this purpose, we have homology-modeled C3-2 using Ig VL and VH domains with a highly similar sequence and a known structure. In the second part, i.e. the actual humanisation step, we have aligned both chains with the most similar mouse and human sequences in order to identify C3-2 amino acid residues that are "typically mouse" and "potentially immunogenic" . For these residues, one or more "candidate human" substitutions were proposed. After verifying the structural compatibility of human candidate residues within the context of the mouse model, as well as their solvent accessibility, a final list of residues-to-be-humanised has been proposed. An alignment of mouse and humanised anti-C3 heavy and light chain variable domains is shown in figure 18.

For eukaryotic expression, both heavy and light chains need to be preceded by suitable signal peptides for correct processing and transport. Therefore, both domains were fused to naturally occuring signal peptide sequences, as obtained from characterisation of human antibodies. The created cleavage site was analysed and predicted to have a high probability for correct cleavage.

The correponding nucleotide sequences for both heavy and light chain humanised variable regions including their signal peptide, consensus Kozak sequences and endonuclease restriction sites for cloning purposes, were made by PCR-based gene assembly and synthesis. By applying standard recombinant DNA methodology, the coding sequence for the humanised VL domain was cloned to a human Ig Kappa light chain constant region resulting in the coding sequence for the complete humanised anti-C3 light chain. The coding sequence for the humanised VH domain was cloned to a human IgGl heavy chain constant region resulting in the coding sequence for the complete humanised anti-C3 heavy chain.

Finally, the coding sequences for the heavy and light chain, were individually introduced into two expression cassettes for expression in eucaryotic cells.

Humanised antibody was produced by a transient transfection of COS-7 cells. Serum-free conditioned medium was harvested at 48 hr and purified using standard chromatographic procedures (protein A affinity chromatography). The humanized C3-2 interacted with native C3 with comparable kinetics as the murine antibody as was shown using BIACORE-analysis (figure 19).

Example 13: Anti-C3-2 antibody prevents human complement-mediated damage of. the rabbit isolated heart.

Materials and Methods

Langendorff preparation: Male NZW rabbits (1.8-2.4 kg) were anaesthetized; after administration of heparin, hearts were excised and flushed with NaCl 0.9% through a catheter in the aorta. Hearts were attached to a modified Langendorff perfusion apparatus, and perfused in a retrograde manner. The perfusion medium consisted of a modified Krebs- Henseleit buffer (pH = 7.44, 36 °C) with additions of BSA (0.25%,w/v) and insulin 10 U/liter. The filtered solution was gassed continuously with a mixture of 95% 0₂/5% CO₂ to achieve the desired oxygen partial pressure of 225 to 300 mm Hg (Micro 13 pH/blood gas analyzer; Instrumentation Laboratory, Lexington, MA). Recalcified human plasma as a source of the complement components was added to the perfusate as described below. Retrograde perfusion was performed at a constant pressure and a recirculating perfusate volume of 500 ml. The hearts were paced via electrodes attached to the right atrium with square wave stimuli from a laboratory stimulator (165 impulses/min, 5 ms duration, 4 V; Grass SD-5, Quincy, MA).

The physiologic parameteres monitored included the aortic flow, the isovolumic left ventricular pressure and its first derivative, dP/dt. The intraventricular fluid-filled latex balloon used for the ventricular pressure measurements was filled to achieve an end-diastolic pressure of 12 to 15 mm Hg. Measurements were recorded continuously (Grass polygraph model 79D), and stored on a digital archive and analysis system (Po-Ne-Mah HD-4, Storrs, CT).

Instrumentation of the heart included cannulation of the pulmonary artery, pulmonary vein ligation, and closure of the left atrial appendage incision around the shunt, thermistor, and ballon tubing to prevent fluid leakage. Treatment protocol: Six treatment groups were used to determine the ability of anti-C3-2 antibody to inhibit the consequences of complement activation on the measured functional parameters in the rabbit isolated perfused heart: 1) control; perfusion with Krebs-Henseleit buffer; 2) control; perfusion with 6% heat-inactivated human plasma; 3) perfusion with 6% normal human plasma; 4) perfusion with 6% normal human plasma in the presence of 3.5 mg anti-C3-2 antibody; 5) perfusion with 6% normal human plasma in the presence of 7.5 mg anti-C3-2 antibody; 6) perfusion with 6% normal human plasma in the presence of 15 mg anti-C3-2 antibody.

Hearts were excised, instrumented and their function was allowed to stabilize for 20 minutes with plasma-free buffer perfusion medium. Base-line functional parameters, as well as the coronary flow were recorded. In treatment groups 4 to 6, anti-C3-2 antibody was added to the reservoir containing the perfusion medium.10 minutes later, 2% (v/v) human plasma was added to the the perfusion medium. Additional human plasma (2% per time point) was added at 15 and 20 minutes after administration of the anti-C3-2 antibody to achieve a final concentration of 6% human plasma in the perfusion medium. The monitored parameters were recorded at 5 minutes intervals through 90 minutes after addition of the anti-C3 -antibody, the end of the protocol. Preparation of normal and heat-inactivated plasma: Human plasma was obtained by venapuncture from fasted donors and frozen. After thawing overnight at 4°C, the plasma was reconstituted with CaCl₂ to a final concentration of 10 mM. Afterwards, the plasma was incubated for 30 minutes at 37° C, to allow a fibrinogen clot to form, which was removed. The plasma was centrifuged (3000 rpm, 15 min) at 4° C; after which it was collected andd stored at -80° C, until shipment on dry ice to the center where the Langendorff experiments were conducted. Heat-inactivated human plasma was prepared by heating (56° C, 1 hour).

Results: Effect of treatment on myocardial function: Hearts perfused with 6% heat-inactivated human plasma served asd the control group. The addition of heat-inactivated human plasma had a moderate effect on the diastolic, systolic, and left ventricular developed pressures (Fig. 20). The contractile function and end-diastolic pressure of hearts perfused with heat-inactivated plasma remained stable through the course of the protocol and exposure to heat-inactived plasma,. Compared to the control group, hearts treated with 6% recalcified human plasma manifested an increase in the left ventricular diastolic pressure, a decrease in the left ventricular systolic pressure, and a concurrent decrease in the left ventricular developed pressure (Fig. 20). Changes in both the systolic and diastolic pressures were evident immediately after the first addition of plasma to the circulating Krebs-Henseleit buffer, and remained like that through the end of the protocol. The inclusion of 15 mg anti-C3-2 antibody protected the hearts from the deleterious effects of perfusion with 6% normal human plasma, with all parameters being at the level of experiments with heat-inactivated plasma, in which all complement has been denatured (Fig 20). Lower doses of anti-C3-2 antibody added to the system were not sufficient to block complement-mediated effects.

References

Aarden La, et al. 1985, lymfokines 10:175

Alper CA, et al. 1969, Science 163 :286 Antibodies: A Laboratory Manual, Harlow et al. Cold Spring Harbor Publications, p. 726

(1988)

Antibody Engineering, Oxford University Press, Oxford (1995) and (1996)

Aoyama et al. 1984, Jpn J Pharmacol 35:203-227

Ausubel et al. (1987) (1992) (1993) Current Protocols in Molecular Biology. Wiley Interscience, N.Y.

Bechere D, et al. 1989, Curr Top Microbiol Immun 153:235

Becherer JD, et al. 1992, Biochemistry 31:1787

Belt KT et al., 1984, Cell 36:907

Campbell RD, Law SKA, Reid KBM, Sim RB, 1988, Annu Rev Immunol; 6:161-195 Campbell, A.M: Monoclonal Antibody Technology Techniques in Biochemistry and

Molecular Biology, Elsevier Science Publishers, Amsterdam, The Netherlands (1984)

Colligan J.E., Kruisbeek A.M., D.H. Margulies D.H. (1992) (1993) Current Protocols in

Immunology, Green Publishing Association and Wiley Interscience, N.Y.

Coloma, M., and S . Morrison. 1997. Nature Biotech. 15:159 Cooper N.R., 1985, Adv Immunol 37:151

De Boer JP, et al. 1993, Infect.hrrmun. 61 :4293

De Brain MHL, et al. 1985, Proc Natl Acad Sci USA 82:708

Fanger M.W.et al. 1992, Crit. Rev. Immunol. 12(3-4):101.

Fishelson Z, 1991, Mol Immunol 28:545 Fujita T. Nature Rev Immunol. 2002 May;2(5): 346-53.

Georgiou G. et al. 1997, Nature Biotech. 15:29.

Glover et al. 1988, Mol. Immunol. 25:1261-1267

Hack CE, et al. 1981, J Immunol Meth 108:77

Hack CE, et al. 1988, J Immunol 141 : 1602 Hack CE, et al. 1988, J Immunol Meth 108:77

Hack CE, et al. 1990, J Immunol 144:4249 Harlow E., David L. et al. (1988) Antibodies: A Laboratory Manual. Cold Spring Harbor

Publications.

Haviland DL, et al. 1991, J Immunol 146:362

Hessing M, et al. 1993, J Immunol Meth 157:39 Hinshaw LB, et al. 1983, J Trauma 23 :361

Hoekzema R, et al. 1985, J Immunol 135:265

Holliger P. et al. 1993, Proc. Natl. Acad. Sci. USA 90(14):6444.

Hoogenboom, H.R. 1997. Trends Biotechnol. 15(2):62.

Hugli TE, 1986 Complement. 1986;3(3):111-27. Hurle MR and Gross M, 1994, Curr. Opin. Biotech 5:428-433.

Ito, W., and Y. Kurosawa. 1993. J. Biol. Chem. 268(27):20668.

Janotova J, 1986, Biochem J 233:819;

Kam et al. 1992, J Immunol 149:163-168.

Kipriyanov S.M. et al. 1996, Protein Eng. 9(2):203. Kirschfink M., 1997, Immunopharmacology 38; 51 -62

Kohler G, et al. 1975, Nature 256:495

Kortt A. et al. 1997, Prot. Eng. 10:423.

Kranz D. et al. 1995, J. Hematother. 4:403.

Lambris JD, 1988, Immunol Today 12:387; Langone J, et al. 1983, Meth Enzymol 92:ρart E.

Law SK, et al. 1981, Biochemistry 20:7457

Luben R, et al. 1980, Mol Immunol 17:635

Lundwall AB et al., 1985, J Biol Chem 260:2108

Lybrand, T.P. 1995. Curr. Opin. Struct. Biol. 5(2):224. Masaki T, et al. 1991, J Immunol 147:927.

Mayforth, Designing Antibodies, Academic Press, San Diego (1993)

McGuinness B. et al. 1996. Nature Biotech. 14:1149.

Mendez et al. 1997, Nat Genet. 15(2): 146-56.

Methods in Molecular Biology, Humana Press, New Jersey (1995). Morris KM, et al. 1982, Science 215:399

MuUer R. (1983) Determination of affinity and specificity of anti-hapten antibodies by competitive radioimirtunoassay. Methods Enzymol. 92: 589-601.

Muller-Eberhard H.J. et al., 1980, Adv Immunol 29:1;

Muller-Eberhard H.J., 1992, In: Gallin JI, Goldstein IM, Snyderman R (eds): Inflammation:

Basic Principles and Clinical Correlates, New York, Raven Press Ltd, p.33. Pack P. et al. 1995, J. Mol. Biol. 246:28.

Panburn M.K., Schreiber R.D., Muller-Eberhard H.J., 1981, J. Exp. Med., vol. 154, 856-867.

Parsons H. et al. 1996, Prot. Eng. 9:1043.

Reading C, 1982, J Immunol Meth 53:261.

Reading C, 1986, Meth Enzymol 121:18; Roguska MA et al. 1994, Proc Natl Acad Sci USA 91:969-973.

Sahu A, Sunyer J.O., Moore W.T., Sarrias M.R., Souhka AM, Lambris J.D., 1998,

Immunologic Research, 1111&2:1Q91121

Sahu et al, 1996, Journal of Immunology, Vol.157, pg 884-891.

Sarubbi E.et al. 1996, Analytical Biochemistry 237:70. Schultz, J. 1996, Biotechnol. Prog. 12:729.

Smeenk RTJ, et al. 1987, Arthr Rheum 30:607

Sottrup- Jensen L et al, 1985, Proc Natl Acad Sci USA 82:9

St. Groth et al. 1980, J. Immunol. Methods 35:1-21

Tack BF, et al. 1976, Biochemistry 15:4513 Tack BF, et al. 1980, Proc Natl Acad Sci USA 77:5764;

Turner MW, 1996, Res Immunol.147(2): 110-5.

Voss B, 1986, Meth Enzymol 121:27

Walport MJ. N Engl J Med. 2001 Apr 12;344(15): 1140-4.

Walport MJ. N Engl J Med. 2001 Apr 5;344(14):1058-66. Wendoloski J.J. et al. 1993, Pharmac. Ther. 60:169.

Wiley, R.A., and D.H. Rich. 1993. Medicinal Research Reviews

Wolbink GJ, et al. 1993, J Immunol Meth 163:67;

Claims

Claims

1. A molecule capable of inhibiting complement activation characterised in that said molecule - specifically binds on a functional domain exposed on native human C3

- thereby inhibiting the generation of the biologically active peptides C3 a and C3b.

2. A molecule according to claim 1, specifically binding on a functional domain exposed on native human C3, wherein said functional domain is homologous to the epitope recognised by the mAb anti-C3-2 produced by the hybridoma deposited according to the Budapest Treaty at the collection of DSMZ under the accession number DSM ACC2562.

3. A molecule according to claim 1-2 wherein said functional domain is in part located on the 23 kD-^α-chain fragment of C3c.

4. A molecule according to claim 1-3 wherein said molecule is an antibody.

5. A molecule according to claim 4 wherein said antibody is a monoclonal antibody.

6. A molecule according to claim 4 or 5 wherein said antibody is a humanised or human antibody.

7. A molecule according to claim 5 wherein said monoclonal antibody is the monoclonal antibody anti-C3-2 produced by the hybridoma deposited according to the Budapest

Treaty at the collection of DSMZ under the accession number DSM ACC2562.

8. A molecule according to claim 1-6 wherein said molecule is competing with the monoclonal antibody anti-C3-2 for binding the functional domain exposed on native C3.

9. Use of a molecule according to claim 1-8 for the preparation of a medicament for inhibiting complement activation.

10. Use of a molecule according to claim 1-8 for the preparation of a medicament for preventing or treating inflammatory diseases mediated by activation of complement.

11. Method for inhibiting complement activation comprising the step of administering a molecule according to claim 1-8.

12. Method for preventing or treating diseases mediated by activation of complement comprising the step of administering a molecule according to claim 1-8.

13. A pharmaceutical composition comprising a molecule according to claim 1-8 in a pharmaceutical acceptable carrier.

14. A method for the preparation of a molecule according to claim 1-8.

15. Use of the mAb anti-C3-2 produced by the hybridoma deposited according to the Budapest Treaty at the collection of DSMZ under the accession number DSM ACC2562, for identification of a molecule according to claim 1-6 or 8.

16. A functional domain exposed on native human C3 characterised in that binding of a molecule according to claim 1-8 on said functional domain prevents the generation of the biologically active peptides C3a and C3b.