Immunology Letters 98 (2005) 49–56

Engineering of human complement component C3 forcatalytic inhibition of complement

Johanna Kolln, Reinhard Bredehorst, Edzard Spillner∗

Institut fur Biochemie und Lebensmittelchemie, Abteilung f¨ur Biochemie und Molekularbiologie, Universit¨at Hamburg,Martin-Luther-King-Platz 6, 20146 Hamburg, Germany

Received 27 August 2004; received in revised form 10 October 2004; accepted 18 October 2004Available online 13 November 2004

Abstract

As a novel therapeutic approach in complement-mediated pathologies, we recently developed a human C3 derivative capable of obliteratingfunctional complement by a catalytic, non-inhibitory mechanism. In this derivative, the C-terminal region of hC3 was substituted by a 275amino acid sequence derived from the corresponding sequence of cobra venom factor (CVF), a complement-activating C3b homologue froms

unogenicitya apable ofo f differentc mplements , its overalli thologies.©

K

1

dartitc[cpcp

om-fore,

lyticase-sma

C3ofase,m-ubse-

net-theolyticd

0d

nake venom.In this study, we replaced shorter C-terminal sequences of hC3 by corresponding CVF sequences to further reduce potential imm

nd to identify domains essential for the formation of functionally stable C3 convertases. In one of these derivatives that is still cbliterating functional complement in vitro, the non-human portion could be reduced to a small domain located in the C-terminus oomplement proteins. This conserved NTR/C345C motif is known to be involved in assembly of different convertases of the coystem. These results suggest a major role of the C345C domain in the regulation of the half-life of the C3 convertase. Moreoverdentity of 96% to human C3 renders this derivative a promising candidate for therapeutic intervention in complement-mediated pa

2004 Elsevier B.V. All rights reserved.

eywords:Complement inactivation; Human complement protein C3; Protein engineering

. Introduction

The complement system is a highly sophisticated host-efense system playing an important role in both innatend antibody-mediated immunity. Pathophysiological dis-egulation of the complement system, however, contributeso tissue damage in many clinical conditions includingschemia-reperfusion injuries after myocardial infarction, in-erventional activation of complement in patients undergoingardiopulmonary bypass, and immune-mediated diseases1–4]. In consequence, there is an urgent need for therapeuticomplement inhibitors. Approaches for inhibition of com-lement developed so far employ the blockade of essentialascade components using inhibitory proteins, small com-ound inhibitors or solubilized receptor molecules[5–7].

∗ Corresponding author. Tel.: +49 40 428386982; fax: +49 40 428387255.E-mail address:spillner@chemie.uni-hamburg.de (E. Spillner).

However, these reagents exert their activity by blocking cplement components at a 1:1 stoichiometry and thereneed to be present in relatively high concentrations.

Recently, we developed an alternative strategy for catainhibition of complement, by modifying the C3 convertof the alternative complement pathway (AP)[8,9]. This pathway is continuously activated at a low rate in human pladue to slow, spontaneous hydrolysis of plasma proteinto C3(H2O) [10]. Upon binding of factor B and cleavageC3b-bound factor B by factor D, the actual AP C3 convertC3bBb, is formed, which initiates amplification of the coplement cascade by increased deposition of C3b and squent formation of new C3 convertases[11]. The natural APC3 convertase, C3bBb, is tightly controlled by a complexwork of regulatory proteins which irreversibly dissociateconvertase subunits and serve as cofactors for the protecleavage of C3b by factor I[12]. Moreover, it is controlleby a half-life limited to less than 2 min[13].

165-2478/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.imlet.2004.10.010

50 J. Kolln et al. / Immunology Letters 98 (2005) 49–56

In order to extend the half-life of this convertase fortherapeutic applications, we substituted structural elementsof human C3 (hC3) by corresponding sequences derivedfrom cobra venom factor (CVF). CVF, a non-toxic 149 kDaC3b homologue from snake venom[14–17], forms aCVF-dependent C3 convertase which escapes regulationby regulatory proteins, is active in fluid phase and exhibitsan increased half-life of 7 h[15]. The CVFBb convertaseobliterates functional complement in vivo for days bycontinuous consumption of complement components[15].CVF has been shown to be highly effective in many exper-imental models of complement-mediated pathologies andhas provided the proof-of-concept for the beneficial effectof catalytic complement depletion (reviewed in[18]). Fortherapeutic applications, however, CVF is not suitable dueto its immunogenic properties.

In our strategy of adopting the mechanism of CVF for ahuman C3 molecule we have replaced in a previous studythe C-terminal region of hC3 by a 275 amino acid sequencederived from the corresponding C-terminal region of CVF[9]. This derivative provides an identity of 91% to human C3and is capable of forming a stable convertase with a half-lifecomparable to that of the CVF-dependent C3 convertase. Fur-thermore, hC3-DIII exhibits significantly reduced C5 conver-tase activity as compared to CVF, thereby reducing harmfuld

ionsi e C3c andr ratest VF-d tableC eser her-a inc

2

2

em(wS ),a oly-c ifiedn

2d

3-d y-p ed as

described[8,9]. For generation of the hC3 derivative DIV-VI,the cDNA of human C3 in pcDNA3 as well as the cDNA ofCVF were used[8]. For the cloning of DIV the fragment ofthe 3′-terminus of the coding sequence of CVF was amplifiedfrom CVF cDNA by PCR using the oligonucleotides S50(TATGTGTACAAAACCAAGCTGCTTCG), AS51 (TTC-TTCTAGATTAAGTAGGGCAGCCAAACTCAGT). Theobtained DNA was inserted into the expression vector ofhC3 in pcDNA3 [8] via the restriction sites ofBsp1407Iand XbaI. Insertion of a N-terminal 6x-his-tag affinity-peptide and an enterokinase cleavage site was performed bygeneration and joining of fragments by PCR using oligo-nucleotides S01 (CTGCTGACTAGTGCGGCCGCTATA-AATATGGGACCCACCTCAGGTCC), AS61 (ATGATG-ATGATGATGATGCCCCAGAGCCAGGGGGAGG), S62(CATCATCATCATCAT-CATGACGATGACGATAAAAG-TCCCAT) and AS03 (AGTACCTTCCGGCTCAGCA-CAACCTCC). The resulting fragment finally was insertedinto the cDNA of hC3 derivative DIV via restriction sitesfor NotI andBpu1102I. For generation of DV two fragmentswere generated using the oligonucleotides S55 (TATGTGTA-CAAGACCCGACTGGTC), AS56 (GGAATTTTTATC-CTTCTCTCCCCAGAAATCGGA) and the cDNA of hC3,as well as the oligonucleotides S57 (TGGGGAGAGAAG-GATAAAAATTCCTACATCATT), AS51 (see above) andt heo r ofhg inga romC G-TA intot s ofB IIa theh

2

dingt ee , Ger-m

2s

Ni-c ny)a urer.T ibed[ onsw wered teri-z

eposition of the anaphylatoxin C5a.In this study, we aimed for refinement of the substitut

n the hC3 molecule required for establishment of stablonvertases to further reduce potential immunogenicityemaining C5 convertase activity. Our study demonsthat a hC3 derivative containing less than 4% of foreign Cerived protein sequences still is capable of forming a s3 convertase with minimal C5 convertase activity. Th

esults emphasize the feasibility of developing novel tpeutics for catalytic reduction of complement activityomplement-mediated pathologies.

. Materials and methods

.1. Materials

Native human C3 was purchased from CalbiochSchwalbach, Germany), native CVF fromNaja kaouthiaas purified according to established protocols[19].treptactin was purchased from IBA (Gottingen, Germanynti-C3 antibody from Cappel (Eschwege, Germany). Plonal anti-CVF-antibodies were raised in goat using purCVF.

.2. Generation of expression plasmids for the hC3erivatives

Generation of full length cDNA of CVF, hC3 and hCerivate DIII providing a N-terminal strep-tag II affiniteptide and an enterokinase cleavage site was perform

he cDNA of CVF. After joining these fragments tbtained DNA was inserted into the expression vectoC3 via the restriction sites ofBsp1407Iand XbaI. Foreneration of hC3 derivative DVI the fragment providstop codon at position 1548 of hC3 was amplified fVF cDNA by PCR using the oligonucleotides S58 (TATGTACAAGACCTGACTGCTTCGAATAGAAGAACAA),S51 (see above). The obtained fragment was inserted

he expression vector of hC3 via the restriction sitesp1407IandXbaI. Insertion of a N-terminal strep-tagffinity-peptide and an enterokinase cleavage site intoC3 derivatives was performed as described[8].

.3. Transfection of CHO cells

Chinese hamster ovary cells (CHO) cultivated accoro standard protocols[19] were transfected with DNA of thxpression vectors using Geneporter (Peqlab, Erlangenany).

.4. Purification of the hC3 derivatives fromupernatant of stably transfected CHO cells

The his-tagged hC3 derivative DIV was purified usinghelat affinity chromatography (Qiagen, Hilden, Germaccording to the recommendations of the manufacthe st-tagged hC3 derivative DIII was purified as descr

9]. Protein concentrations and purity of the fractiere analyzed by 7.5% SDS-PAGE. Pooled fractionsialyzed against PBS and employed for further characation.

J. Kolln et al. / Immunology Letters 98 (2005) 49–56 51

2.5. Complement consumption assay

The complement-consuming activity was determined ac-cording to established protocols[20] with slight modifica-tions. Briefly, 20�l of protein sample or GVBS++ (2.5 mMNa-5,5-diethyl-barbituric acid, 143 mM NaCl, 0.75 mMMgCl2, 0.15 mM CaCl2, 0.1% Gelatine, pH 7.4; serumcontrol) was mixed with 20�l normal human serum (di-luted 1:2 in GVBS++) and incubated for 3 h at 37◦C. Af-ter addition of 100�l of GVBS++ and 30�l of sensitizedsheep erythrocytes (5× 108 cells/ml) generated by incuba-tion with anti-sheep erythrocyte antibodies, the incubationwas continued at 37◦C. Every 10 min of incubation, con-trol samples containing no serum (GVBS++ control) or onlyserum (serum control) were supplemented with 850�l ice-cold GVBS++, centrifuged and analyzed for the release ofhemoglobin. When serum control values reached approxi-mately 80% of maximal lysis, all samples were processed asdescribed above. GVBS++ control containing only erythro-cytes was performed parallel. The percentage of lysis wasdetermined according to formula (sample− GVBS++ con-trol)/(serum control− GVBS++ control)× 100.

2.6. Solid phase complement consumption assay ofstrep-tag II-derivatized proteins

iTt /v).A dif-f sionsw at4 ithPc nsi-t wasi 50r -m d into23 mentc

2

C5c shedpB2 s( sc aterc oldG zeda mined

according to formula (sample− serum control)/(water con-trol − serum control)× 100.

2.8. Other methods

SDS-PAGE, western blotting as well as standard proce-dures in molecular biology were performed according to es-tablished protocols[21].

3. Results

3.1. Generation and transient expression of hC3derivatives

Utilizing the framework of hC3 and C-terminal structuralelements of CVF, we generated several hC3 derivatives.The previously reported biologically active hC3 derivative(hC3-DIII) in which the C-terminal 275 amino acid residueshad been replaced by a corresponding CVF-derived sequence[9], provided the basis for the design. In order to identifythose amino acid residues in the C-terminus of hC3 whichneed to be replaced by corresponding CVF-derived residuesfor the generation of biologically active hC3 derivatives,we first compared the primary structure of both the hC3-a encea int 275a morer 275a tieso threea

lyt rag-m en-t tot 116a Thec ntityt yseso as-s oc-t ers

uentt le int blot-t ctswa tions,r ticala ilart C-t d as

Streptactin (3�g in 40�l 0.1 M NaHCO3, pH 9.5) wasmmobilized onto polystyrene supports overnight at 4◦C.hereafter, the wells were rinsed 3 times with 200�l PBS-

ween 20 (0.1%, v/v) and blocked with PBS-BSA (3%, wfter washing 3 times with PBS-tween 20 (0.1%, v/v),

erent volumes of the supernatants of transient expresere applied to the wells. After incubation overnight◦C under agitation, the wells were rinsed 3 times wBS-tween 20 (0.1%, v/v). Subsequently, 60�l GVBS++

ontaining normal human serum for 80% lysis of seized erythrocytes were added and the ELISA platencubated for 3 h at 37◦C under vigorous shaking at 1pm (InnovaTM, New Brunswick Scientific, Nurtingen, Gerany). Subsequently, the supernatants were transferreml reaction tubes, mixed with 100�l of GVBS++ and0�l of sensitized erythrocytes, and assayed for compleonsumption.

.7. Bystander lysis assay

The Bystander lysis assay for analysis of fluid phaseonvertase activity was carried out according to establirotocols[19] with slight modifications as described[9].riefly, the sample (20�l) was mixed with 20�l GVBS++,0�l guinea pig serum and 20�l of guinea pig erythrocyte5× 108 cells/ml) and incubated for 3 h at 37◦C. Sampleontaining no protein (serum control) or no serum (wontrol) were used as controls. After addition of 1 ml ice-cVBS++ or water, respectively, the samples were analys described above. The percentage of lysis was deter

nd the CVF-derived 275 amino acids. The sequlignment shown inFig. 1, revealed more homology

he N-terminal than in the C-terminal regions of themino acid fragments. This observation suggest aelevant contribution of the C-terminal sequences of themino acid fragments to the different biological properf hC3 and CVF. To test this hypothesis, we generateddditional hC3 derivatives.

As shown inFig. 2, in construct DIV we substituted onhe 116 C-terminal amino acids of the 275 amino acid fent. This 116 amino acid fragment exhibits only 44% id

ity. In the second construct DV, substitution was limitedhe C-terminal 48 amino acids. As control, the terminalmino acids of the hC3 were eliminated in construct DVI.onstructs DIV and DV display an overall sequence ideo human C3 of 96% and 98%, respectively. For analf the complement-consuming activity in a solid phaseay all hC3 derivatives were derivatized with the affinityapeptide strep-tag II (St)[22] at the C-terminus of the leadequence.

After cloning of the expression plasmids and subseqransfection of CHO cells, the derivatives were detectabhe culture supernatants by SDS-PAGE and Westerning (Fig. 3A). The expression levels of these construere found to be in the range of 1–2�g/ml. In SDS-PAGEnd western blotting analyses under non-reducing condiCVF, hC3, and the hC3 derivatives exhibited an idenpparent molecular weight in the range of 200 kDa, sim

o that of native hC3. According to the deletion of theerminal 116 amino acids the hC3 derivate DVI exhibitelightly reduced molecular weight.

52 J. Kolln et al. / Immunology Letters 98 (2005) 49–56

Fig. 1. Sequence alignment of the C-terminal domains of human C3 and CVF. Comparison of the 275 C-terminal amino acids of hC3 and CVF. Alignmentwas performed using clustalW. Disulfide bonds are shown as lines connecting cystein residues (in squares). The residues substituted in hC3-DIV or eliminatedin DVI are bold, the sequences substituted in hC3-DV is underlined. Identity is indicated by asterisk, conservative or semi-conservative substitutions by two orone dots. The putative C345C domains of hC3 and CVF are shaded gray.

The proteins in supernatants of transiently transfectedcells were assayed for complement consumption activity bya solid phase assay employing streptactin-coated supports[8]. Based on the intensity of the western blot bands, theconcentrations of the recombinant protein were adjustedto comparable values. Immobilization of strep-tag II-derivatized proteins to the support utilized for the solid phaseassay was verified by ELISA-techniques (Fig. 3B). Deriva-tive hC3-DIII and rCVF were used as control. The derivativehC3-DVI exhibited slight complement consuming activity,whereas derivatives DV and DVI did not effect any detectableconsumption of complement (Fig. 3C). These results couldbe verified by a fluid phase assay utilizing the recombinant

proteins after immunoprecipitation from supernatants oftransiently transfected cells as described[8] (data not shown).

3.2. Functional analysis of purified hC3 derivative DIV

To analyze the functional activity of derivative DIV morein detail, the protein was modified with a 6xhis-tag at theC-terminus of the leader sequence. After establishment ofa stably transfected cell line, the recombinant protein wasenriched from cellular supernatants using Ni-chelate affinitychromatography. Analysis by SDS-PAGE and Western blot-ting under non-reducing and reducing conditions revealed anidentical molecular mass as the strep tag II-derivatized pro-

F of hC3 nd DVI.I s well

ig. 2. Design of hC3, CVF and hC3 derivatives. Primary structuresndicated are the sites for binding or cleavage of regulatory proteins a

(dark gray), CVF (light gray) and the hC3 derivatives DIII, DIV, DV aas nomenclature of the chains of the resulting proteins.

J. Kolln et al. / Immunology Letters 98 (2005) 49–56 53

Fig. 3. Transient expression of hC3, CVF and hC3 derivatives. A: Westernblot analysis of the transiently expressed proteins. Native hC3 and culturesupernatants of transiently transfected CHO cells containing rCVF, rhC3and the hC3 derivatives were separated by SDS-PAGE (7.5%) under non-reducing conditions, transferred to a PVDF-membrane, and detected withpolyclonal anti-CVF antibodies (rCVF) or anti-hC3 antibodies (hC3, rhC3,derivatives) as described in Section2. B: Immobilisation of transiently strep-tag II-derivatized proteins onto streptactin-coated supports was analysed byELISA. Detection of the bound proteins was performed using anti-CVFor anti-hC3 polyclonal antibodies from goat, diluted 1:1000 in PBS-BSA(1.5%, w/v), and anti-goat HRP-conjugate, diluted 1:1000 in PBS-BSA(1.5%, w/v). Controls (open bars) were performed by omission of streptactin.C: Solid-phase complement consumption assay. Strep-tag II-derivatizedproteins were immobilized to streptactin-coated supports and assayed forcomplement consuming activity as described in Section2. Strep-tag II-derivatized hC3 was utilized as negative control. Data inFig. 2 representmean values± S.D. obtained from at least three independent experiments.

teins, and a two chain structure as demonstrated for hC3 andrCVF expressed in mammalian cells (not shown)[8].

As shown inFig. 4 the enriched derivative DIV was ana-lyzed for complement-consuming activity over a broad con-centration range. The obtained data confirmed the CVF-likeactivity observed in analyses of transiently expressed hC3derivative. The hC3 derivative DIV exhibited a complement-consuming activity which is comparable to that of DIII andapproximately 75% of that of purified native CVF.

Fig. 4. Characterization of stably expressed derivative DIV. A: Complementconsumption assay. Purified hC3 (gray dots), nCVF (gray squares), hC3-DIII(black up triangles), or hC3-DIV (black down triangles) were assayed forcomplement-consuming activity as described in Section2. B: Bystanderlysis assay. Purified hC3 (gray dots), nCVF (gray squares), hC3-DIII (blackup triangles), and hC3-DIV (black down triangles) were assayed for fluidphase C5 convertase activity utilizing bystander lysis assay as described inSection2.

Since CVFBb also exerts C5 convertase activity in fluidphase[23], derivative hC3-DIV was analyzed for potentialfluid phase C5 convertase activity in a bystander lysis assay(Fig. 4B). The capability of derivative DIV for activation ofguinea pig serum in fluid phase was significantly reduced incomparison to CVF. Only 15% of the fluid phase C5 conver-tase activity of CVF was found to be retained in the deriva-tive, which is slightly lower than the C5 convertase activityretained in derivative DIII.

4. Discussion

Recently, we developed a human C3 derivative capableof forming stable C3 convertases in human serum with ahalf-life in the range of 5–6 h similar to that of CVF[9]. Thisderivative provided evidence for the fact that substitutionof the C-terminal region in the hC3 molecule by thecorresponding CVF sequence is sufficient for conferring aninherent molecular stability to the C3-dependent convertasecomplex. Although this derivative, hC3-DIII, exhibits a

54 J. Kolln et al. / Immunology Letters 98 (2005) 49–56

homology of approximately 91% to hC3, for therapeutic ap-proaches a further reduction of non-human sequences wouldbe desirable to minimize the potential immunogenicity.Furthermore, the CVF-derived sequence conferred also fluidphase C5 convertase activity, that is exerted exclusively bythe CVF-dependent, but not by the C3-dependent convertase[23], to derivative DIII in the range of 20%. Nevertheless,a further reduction or even elimination would be favorablesince the deposition of anaphylatoxin C5a is known to causesevere tissue damage[24,25].

Although in hC3-DIV a 116 amino acid sequence has beenreplaced by the corresponding CVF-derived sequence, only66 out of 1653 amino acids are of foreign origin, correspond-ing to a sequence identity of 96%. Accordingly, the immuno-genicity of the hC3 derivative can be assumed to be verylow to negligible. Therapeutic antibodies generally providebetween 65 and 95% human origin[26–28], and even the ad-ministration of an anti-C5 scFv containing murine CDR andartificial linker sequences was found to be well tolerated anddid not provoke any immune response[29,30]. Furthermore,the high identity of hC3-DIV to the high abundance serumprotein hC3 provides additional important advantages withregard to pharmacokinetics and safety. The short C-terminalsequence derived from the non-toxic CVF[31,32], is unlikelyto affect neither the safety nor the biochemical characteristicso es-s medt im-pD ec tionr l hC3d ck-i ed tob e an-i aleda tudel eh tra-t

t in-h wn.T C5aa d byr Fur-t asea er-t lowert n-c ctiona elyt ces-s haves ientt ritis[ er

reduction of the non-human sequences resulted in inactivemolecules as revealed by complement consumption analysesof supernatants of transient expressions. Apparently, a certainstretch of the C-terminal domain of CVF is essential for thegeneration of an architecture which provides tight bindingof Bb. Although the precise molecular mechanism remainsto be elucidated, sequence analysis revealed the presence ofa C345 domain in the C-terminal 116 amino acids of CVFas shown inFig. 1. The N-terminus of this domain includes15 additional amino acids, however, since 11 of these aminoacids are identical and further two replaced conservatively,only 98 of the 113 amino acids have been substituted. Thecorresponding C-terminal stretch of C3 is known to containthis conserved C345C or NTR domain[38]. This domain isalso present in the C-terminus of C4 and C5, and homolo-gous modules have been found in other proteins includingnetrins and tissue inhibitors of metalloproteinases (TIMPs)throughout a variety of species[39–42]. As revealed by struc-tural analyses the C345C module is characterized by a highcontent of beta-sheets, which in C3 and C5 additionally arestabilized by three internal disulfide bonds[43]. The impor-tance of amino acid residues within the C345C domain of C5for proteolytic activation by the classical pathway C5 con-vertase[44,45] suggests that these residues in C5–C345Ccontribute to the contact between C5 and the convertase. Po-t tiono apa-b sesa f theC ain-i aili C3-d einer eralls 45Cd dif-f 345Cd em

A

them k P.Z

R

ylortion

rthri-Im-

E,pul-

f the molecule. Utilizing mammalian cells for the exprion of hC3-DIV the glycosylation pattern can be assuo be similar to that of hC3. Since glycosylation plays anortant role for the half-life of serum proteins[33,34], hC3-IV is likely to provide a half-life sufficient for effectivomplement depletion in vivo. The catalytic mode of acepresents the most important advantage of the noveerivative. Currently available complement inhibitors, blo

ng defined components of the complement cascade, nee administered in stoichiometric amounts. Comparativ

mal studies with CVF and selected blocking agents reven effective dose of CVF that is 1 to 2 orders of magni

ower than those of blocking agents[35,36]. Therefore, thC3 derivative probably will require much lower concen

ions for effective decomplementation.The exact risks associated with catalytical complemen

ibition in man via the novel hC3 derivatives are not knohe generation of the anaphylatoxic peptides C3a andccompanying complement activation could be controlleepeated injections of low doses of these derivatives.hermore, derivative DIV exhibits very low C5 convertctivity, and the relative biological activity of C3a for c

ain responses has been reported to be at least 10-foldhan that of C5a[25]. It is known, that complement deficieies are associated with increased susceptibility to infend with autoimmunity. Such complications would be lik

o arise only when total complement inhibition was neary to achieve therapeutical benefit. However, studieshown that even 60% inhibition of complement is suffico provide therapeutic benefit in collagen-induced arth37]. In the additional derivatives DV and DVI the furth

entially, this domain may contribute also to the stabilizaf the C3 convertases of the alternative pathway. The cility of the hC3 derivatives of forming of stable convertapparently is associated with the maintained integrity oVF-derived C345C module, while the constructs cont

ng a hybrid module (DV) or no C345C domain (DVI) fn forming a stable convertase. Interestingly, in both theerived as well as the CVF-derived module the 6 cystesidues for disulfide-bridging are retained and the ovtructure may be similar. Recombinant expression of C3omains of CVF and hC3 for interaction analysis with

erent complement components as performed for the Comains of C5 and C3[46] may help to further dissect thechanism of convertase stabilization.

cknowledgements

We gratefully acknowledge critical reading ofanuscript by K. Greunke and I. Braren and thaniegelmuller for providing purified nCVF.

eferences

[1] Kemp PA, Spragg JH, Brown JC, Morgan BP, Gunn CA, TaPW. Immunohistochemical determination of complement activain joint tissues of patients with rheumatoid arthritis and osteoatis using neoantigen-specific monoclonal antibodies. J Clin Labmunol 1992;37:147–62.

[2] Kirklin JK, Westaby S, Blackstone EH, Kirklin JW, Chenoweth DPacifico AD. Complement and the damaging effects of cardiomonary bypass. J Thorac Cardiovasc Surg 1983;86:845–57.

J. Kolln et al. / Immunology Letters 98 (2005) 49–56 55

[3] Homeister JW, Satoh P, Lucchesi BR. Effects of complement acti-vation in the isolated heart. Role of the terminal complement com-ponents. Circ Res 1992;71:303–19.

[4] Hill JA, Ward PA. The phlogistic role of C3 leukotactic fragmentsin myocardial infarcts of rats. J Exp Med 1971;133:885–900.

[5] Weisman HF, Bartow T, Leppo MK, Marsh Jr HC, Carson GR, Con-cino MF, Boyle MP, Roux KH, Weisfeldt ML, Fearon DT. Solublehuman complement receptor type 1: in vivo inhibitor of comple-ment suppressing post-ischemic myocardial inflammation and necro-sis. Science 1990;249:146–51.

[6] Evans MJ, Rollins SA, Wolff DW, Rother RP, Norin AJ, TherrienDM, Grijalva GA, Mueller JP, Nye SH, Squinto SP, Wilkins JA. Invitro and in vivo inhibition of complement activity by a single-chainFv fragment recognizing human C5. Mol Immunol 1995;32:1183–95.

[7] Fiane AE, Mollnes TE, Videm V, Hovig T, Hogasen K, MellbyeOJ, Spruce L, Moore WT, Sahu A, Lambris JD. Prolongation ofex vivo-perfused pig xenograft survival by the complement inhibitorcompstatin. Transplant Proc 1999;31:934–5.

[8] Kolln J, Matzas M, Janner N, Mix T, Klensang K, Bredehorst R,Spillner E. Functional analysis of cobra venom factor/human C3chimeras transiently expressed in mammalian cells. Mol Immunol2004;41:19–28.

[9] Kolln J, Spillner E, Andra J, Klensang K, Bredehorst R. Comple-ment inactivation by recombinant human C3 derivatives. J Immunol2004;173(3):5540–5.

[10] Pangburn MK, Muller-Eberhard HJ. The alternative pathway of com-plement. Springer Semin Immunopathol 1984;7:163–92.

[11] Fearon DT, Austen KF. Properdin: binding to C3b and stabilizationof the C3b-dependent C3 convertase. J Exp Med 1975;142:856–63.

[ rs.York:

[ aybileMed

[ en-unol

[ dentamiceight

[the

Biol

[ vedS A

[ rrent

[ vedunol

[ bram. J

[ ew

[ tionstrep-

[ c-e C5

[24] Schmid E, Warner RL, Crouch LD, Friedl HP, Till GO, HugliTE, Ward PA. Neutrophil chemotactic activity and C5a fol-lowing systemic activation of complement in rats. Inflammation1997;21:325–33.

[25] Wetsel RA, Kolb WP. Expression of C5a-like biological activities bythe fifth component of human complement (C5) upon limited diges-tion with noncomplement enzymes without release of polypeptidefragments. J Exp Med 1983;157:2029–48.

[26] Bruggemann M, Winter G, Waldmann H, Neuberger MS. The im-munogenicity of chimeric antibodies. J Exp Med 1989;170:2153–7.

[27] Riechmann L, Clark M, Waldmann H, Winter G. Reshaping humanantibodies for therapy. Nature 1988;332:323–7.

[28] Hale G, Dyer MJ, Clark MR, Phillips JM, Marcus R, RiechmannL, Winter G, Waldmann H. Remission induction in non-Hodgkinlymphoma with reshaped human monoclonal antibody CAMPATH-1H. Lancet 1988;2:1394–9.

[29] Thomas TC, Rollins SA, Rother RP, Giannoni MA, Hartman SL,Elliott EA, Nye SH, Matis LA, Squinto SP, Evans MJ. Inhibitionof complement activity by humanized anti-C5 antibody and single-chain Fv. Mol Immunol 1997;33:1389–401.

[30] Fitch JC, Rollins S, Matis L, Alford B, Aranki S, Collard CD,Dewar M, Elefteriades J, Hines R, Kopf G, Kraker P, Li L, O’HaraR, Rinder C, Rinder H, Shaw R, Smith B, Stahl G, Shernan SK.Circulation 1999;100:2499–506.

[31] Cochrane CG, Muller-Eberhard HJ, Aikin BS. Depletion of plasmacomplement in vivo by a protein of cobra venom: its effect on variousimmunologic reactions. J Immunol 1970;105:55–69.

[32] Vogel CW. In: Tu AT, editor. Handbook of Natural Toxins, vol. 5.New York: Marcel Dekker Inc.; 1991. p. 147–88.

[33] Wright A, Morrison SL. Effect of C2-associated carbohydrate struc-manvary

[ tosyllls. J

[ SM,FH.entgraft

[ cep-tment–98.

[ alrates

9.[ en-

[ etedancerses.

[ n-bout–44.

[ acherin-iol

[ ngypesus ofctivity

[ ment

12] Liszewski MK, Atkinson JP. In: Volanakis JE, Frank MM, editoThe Human Complement System in Health and Disease. NewMarcel Dekker Inc.; 1998. p. 149–66.

13] Medicus RG, Gotze O, Muller-Eberhard HJ. Alternative pathwof complement: recruitment of precursor properdin by the laC3/C5 convertase and the potentiation of the pathway. J Exp1976;144:1076–93.

14] Muller-Eberhard HJ, Fjellstrom KE. Isolation of the anticomplemtary protein from cobra venom, its mode of action on C3. J Imm1971;107:1666–72.

15] Vogel CW, Muller-Eberhard HJ. The cobra venom factor-depenC3 convertase of human complement. A kinetic and thermodynanalysis of a protease acting on its natural high molecular wsubstrate. J Biol Chem 1982;257:8292–9.

16] Vogel CW, Bredehorst R, Fritzinger DC, Grunwald T, ZiegelmullerP, Kock MA. Structure and function of cobra venom factor,complement-activating protein in cobra venom. Adv Exp Med1996;391:97–114.

17] Fritzinger DC, Bredehorst R, Vogel CW. Molecular cloning, deriprimary structure of cobra venom factor. Proc Natl Acad Sci U1994;91:12775–9.

18] Morgan BP, Harris CL. Complement therapeutics; history and cuprogress. Mol Immunol 2003;40:159–70.

19] Vogel CW, Muller-Eberhard HJ. Cobra venom factor: impromethod for purification and biochemical characterization. J ImmMeth 1984;73:203–20.

20] Ballow M, Cochrane CG. Two anticomplementary factors in covenom: hemolysis of guinea pig erythrocytes by one of theImmunol 1969;103:944–52.

21] Ausubel FM, editor. Current Protocols in Molecular Biology. NYork: Wiley Interscience; 1996.

22] Schmidt TG, Koepke J, Frank R, Skerra A. Molecular interacbetween the Strep-tag affinity peptide and its cognate target,tavidin. J Mol Biol 1996;255:753–66.

23] DiScipio RG, Smith CA, Muller-Eberhard HJ, Hugli TE. The ativation of human complement component C5 by a fluid phasconvertase. J Biol Chem 1983;258:10629–36.

ture on Ig effector function: studies with chimeric mouse-huIgG1 antibodies in glycosylation mutants of Chinese hamster ocells. J Immunol 1998;1607:3393–402.

34] Dong X, Storkus WJ, Salter RD. Binding and uptake of agalacIgG by mannose receptor on macrophages and dendritic ceImmunol 1999;16310:5427–34.

35] Candinas D, Lesnikoski BA, Robson SC, Miyatake T, ScesneyMarsh Jr HC, Ryan US, Dalmasso AP, Hancock WW, BachEffect of repetitive high-dose treatment with soluble complemreceptor type 1 and cobra venom factor on discordant xenosurvival. Transplantation 1996;15:336–42.

36] Vriesendorp FJ, Flynn RE, Pappolla MA. Soluble complement retor 1 (sCR1) is not as effective as cobra venom factor in the treaof experimental allergic neuritis. Intern J Neurosci 1997;92:287

37] Wang Y, Rollins SA, Madri JA, Matis LA. Anti-C5 monoclonantibody therapy prevents collagen-induced arthritis and amelioestablished disease. Proc Natl Acad SciU S A 1995;92(19):8955–

38] Bork P, Bairoch A. Extracellular protein modules: a proposed nomclature. Trends Biochem Sci 1995;20. Poster suppl.

39] Banyai L, Patthy L. The NTR module: domains of netrins, secrfrizzled related proteins, type I procollagen C-proteinase enhprotein are homologous with tissue inhibitors of metalloproteinaProtein Sci 1999;8:1636–42.

40] Chong JM, Rubin JS,Uren A, Speicher DW. Disulfide bond assigments of secreted frizzled-related protein-1 provide insights afrizzled homology, netrin modules. J Biol Chem 2002;277:5134

41] Stetefeld J, Jenny M, Schulthess T, Landwehr R, SchumB, Frank S, Ruegg MA, Engel J, Kammerer RA. The laminbinding domain of agrin is related to N-TIMP-1. Nat Struct B2001;8:705–9.

42] Liepinsh E, Banyai L, Pintacuda G, Trexler M, Patthy L, OttiG. NMR structure of the netrin-like domain (NTR) of human tI procollagen C-proteinase enhancer defines structural consenNTR domains and assesses potential proteinase inhibitory aand ligand binding. J Biol Chem 2003;278:25982–9.

43] Dolmer K, Sottrup-Jensen L. Disulfide bridges in human complecomponent C3b. FEBS Lett 1993;315:85–90.

56 J. Kolln et al. / Immunology Letters 98 (2005) 49–56

[44] Low PJ, Ai R, Ogata RT. Active sites in complement componentsC5 and C3 identified by proximity to indels in the C3/4/5 proteinfamily. J Immunol 1999;162:6580–8.

[45] Sandoval A, Ai R, Ostresh JM, Ogata RT. Distal recognition sitefor classical pathway convertase located in the C345C/netrin mod-

ule of complement component C5. J Immunol 2000;165:1066–73.

[46] Thai M, Ogata RT. Expression and characterization of theC345C/NTR domains of complement components C3 and C5. J Im-munol 2003;171(12):6565–73.