crystal structure of the cub1 -egf-cub2 domain of human masp

Crystal Structure of the CUB1-EGF-CUB2 Domain of HumanMASP-1/3 and Identification of Its Interaction Sites withMannan-binding Lectin and Ficolins*

Received for publication, May 9, 2008, and in revised form, June 11, 2008 Published, JBC Papers in Press, July 2, 2008, DOI 10.1074/jbc.M803551200

Florence Teillet‡1, Christine Gaboriaud§1, Monique Lacroix‡, Lydie Martin§, Gerard J. Arlaud‡,and Nicole M. Thielens‡2

From the ‡Laboratoire d’Enzymologie Moleculaire and the §Laboratoire de Cristallographie et Cristallogenese des Proteines,Institut de Biologie Structurale Jean-Pierre Ebel, CNRS-CEA-UJF, UMR 5075, 41 Rue Jules Horowitz, 38027 Grenoble Cedex 1, France

MASP-1 andMASP-3 are homologous proteases arising fromalternative splicing of theMASP1/3 gene. They include an iden-tical CUB1-EGF-CUB2-CCP1-CCP2 module array prolonged bydifferent serine protease domains at the C-terminal end. Thex-ray structure of the CUB1-EGF-CUB2 domain of humanMASP-1/3, responsible for interaction of MASP-1 and -3 withtheir partner proteins mannan-binding lectin (MBL) and fico-lins, was solved to a resolution of 2.3 A. The structure shows ahead-to-tail homodimer mainly stabilized by hydrophobicinteractions between theCUB1module of onemonomer and theepidermal growth factor (EGF) module of its counterpart. ACa2� ion bound primarily to both EGF modules stabilizes theintra- and inter-monomer CUB1-EGF interfaces. AdditionalCa2� ions are bound to each CUB1 and CUB2 module throughsix ligands contributed by Glu49, Asp57, Asp102, and Ser104(CUB1) and their counterpartsGlu216,Asp226,Asp263, andSer265(CUB2), plus one and two water molecules, respectively. Toidentify the residues involved in interaction of MASP-1 and -3with MBL and L- and H-ficolins, 27 point mutants of humanMASP-3were generated, and their binding propertieswere ana-lyzed using surface plasmon resonance spectroscopy. Thesemutations map two homologous binding sites contributed bymodules CUB1 andCUB2, located in close vicinity of their Ca2�-binding sites and stabilized by the Ca2� ion. This informationallows us to propose amodel of theMBL-MASP-1/3 interaction,involving a major electrostatic interaction between two acidicCa2� ligands ofMASP-1/3 and a conserved lysine ofMBL.Basedon these and other data, a schematic model of a MBL�MASPcomplex is proposed.

The lectin pathway of complement is increasingly recog-nized as an important component of innate immunity againstpathogens. This pathway is triggered by oligomeric lectins that

recognize patterns of neutral and acetylated carbohydrates onthe surface of pathogens and share the ability to associate withand trigger activation of modular proteases termed mannan-binding lectin-associated serine proteases (MASPs)3 (1, 2).Four such oligomeric lectins have been described: mannan-binding lectin (MBL) and ficolins L, H, and M (3–7). Theseproteins all assemble as oligomers of homotrimeric subunits,each comprising N-terminal collagen-like triple helices pro-longed by recognition domains endowed with lectin-like bind-ing activities. There are three different MASPs (MASP-1, -2,and -3) (4, 8, 9), and these feature modular structures homolo-gous to those ofC1r andC1s, the proteases of theC1 complex ofcomplement, with an N-terminal CUBmodule (10), an epider-mal growth factor (EGF)-like module belonging to the Ca2�-binding subset (11), a second CUB module, two complementcontrol protein (CCP) modules (12), and a chymotrypsin-likeserine protease domain. MASP-1 and MASP-3 are alternativesplicing products of the MASP1/3 gene and include differentserine protease domains but share identical CUB1-EGF-CUB2-CCP1-CCP2 segments (9). Likewise, alternative splicing of theMASP2 gene generates MBL-associated protein 19 (MAp19), atruncated protein comprising the N-terminal CUB1-EGF seg-ment ofMASP-2 prolonged by four residues specific toMAp19(13, 14). From a functional standpoint, the ability ofMASP-2 totrigger the lectin pathway of complement is clearly established(8). In contrast, whether MASP-1 is involved in this pathway isstill a controversial issue (15, 16), and the function of MASP-3remains elusive.Studies on human (17–20) and rat (21, 22) proteins have

established that the MASPs and MAp19 each associate ashomodimers through their CUB1-EGF segment. In turn, theMASPs and MAp19 each form individual Ca2�-dependentcomplexes with MBL and the ficolins. The interaction involvesa major site located in the CUB1-EGF moiety of each proteinbut is strengthened bymodule CUB2 (18, 19, 22, 23). Resolutionof the x-ray structure of humanMAp19 has allowed identifica-tion of a Ca2�-binding site in module CUB1, and site-directedmutagenesis has provided evidence that this underlies the bind-ing site for MBL and L-ficolin (20). The x-ray structure of the

* This work was supported by the Commissariat a l’Energie Atomique, theCNRS, and the Universite Joseph Fourier, Grenoble, France. The costs ofpublication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement” inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The atomic coordinates and structure factors (code 3DEM) have been depositedin the Protein Data Bank, Research Collaboratory for Structural Bioinformat-ics, Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

1 Both authors contributed equally to this work.2 To whom correspondence should be addressed. Tel.: 33 4 38 78 95 79; Fax:

33 4 38 78 54 94; E-mail: [email protected].

3 The abbreviations used are: MASP, mannan-binding lectin-associated ser-ine protease; CCP, complement control protein; CUB, protein module orig-inally identified in complement proteins C1r/C1s, Uegf, and bone morpho-genetic protein; EGF, epidermal growth factor; MBL, mannan-bindinglectin; SP, serine protease; PEG, polyethylene glycol.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 283, NO. 37, pp. 25715–25724, September 12, 2008© 2008 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

SEPTEMBER 12, 2008 • VOLUME 283 • NUMBER 37 JOURNAL OF BIOLOGICAL CHEMISTRY 25715

by guest on April 8, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

CUB1-EGF-CUB2 segment of rat MASP-2 was also solved, butin contrast no Ca2� ion could be observed in the CUBmodules,and therefore a different model for interaction with MBL wasproposed (22). On the other hand, expression of site-directedmutants of human and ratMBL (24, 25) and ficolins (26, 27) hasrecently provided evidence that these proteins associate withthe MASPs and MAp19 through a major ionic interactioninvolving a conserved lysine residue.We now report the crystal structure of the CUB1-EGF-CUB2

domain of human MASP-1/3. The homodimeric structure issimilar to that described for the homologous rat MASP-2domain but holds a Ca2�-binding site in each CUB1 and CUB2module. In line with our previous studies on MAp19 (20), weprovide evidence that each of these sites is involved in the inter-action with MBL and the ficolins and propose a schematicmodel of a MBL�MASP complex.

EXPERIMENTAL PROCEDURES

Proteins—MBLwas purified fromhuman serum as describedby Zundel et al. (19). The CUB1-EGF-CUB2 segment of humanMASP-1/3 was expressed in a baculovirus/insect cell systemand purified as described previously (18). Full-length wild-typeMASP-3 and its variants were expressed in a baculovirus/insectcell system (19). All mutants were secreted in the culturemedium, with yields of 8–10 �g/ml, similar to those obtainedfor the wild-type form, except E216A and D226A, which wereproduced at 1 and 5 �g/ml, respectively. All MASP-3 variantswere purified by ion-exchange chromatography on a Q-Sepha-rose Fast Flow column (GE Healthcare) followed by high pres-sure gel permeation on a TSK G3000 SWG column (TosoHaas), as described previously (25). This latter step allowed usto check that all MASP-3 variants retained the ability to asso-ciate as homodimers. The concentrations of purified proteinswere determined using absorption coefficients (A1%, 1 cm at280 nm) calculated by the method of Gill and von Hippel (28)and molecular weights determined by mass spectrometry, asfollows: MASP-1/3 CUB1-EGF-CUB2, 10.0 and 34,200 (18);MBL, 7.6 and 25,340 (29); L-ficolin, 17.6 and 33,800; H-ficolin,19.6 and 32,800; MASP-3 variants, 13.5 and 87,500, exceptY56A and Y225A (13.3 and 87,500).Expression and Purification of H-ficolin—A DNA segment

encoding the H-ficolin signal peptide plus the mature protein(amino acid residues 1–276) was amplified by PCR using thepGEM-T easy plasmid containing the full-length cDNA (30) asa template, according to established procedures. The amplifiedH-ficolin DNA was cloned into the pcDNA3.1(�) mammalianexpression plasmid (Invitrogen). A stable CHO K1 cell lineexpressingH-ficolin was created usingGeneticin (G418 sulfate,Invitrogen), and the recombinant protein was produced inserum-free medium as described by Teillet et al. (25). Recom-binant H-ficolin was purified from the culture supernatant byion-exchange chromatography on a Q-Sepharose Fast Flowcolumn (GE Healthcare) equilibrated in 5 mM CaCl2, 50 mM

triethanolamine-HCl, pH 8.0, using a linear NaCl gradient to250 mM. Fractions containing the recombinant protein wereidentified by Western blot analysis, dialyzed against 145 mM

NaCl, 5 mM CaCl2, 20 mM Tris-HCl, pH 7.4, and further puri-

fied by gel filtration on a Superose 6 column (GE Healthcare)equilibrated in the same buffer.Purification of L-ficolin—L-ficolin was purified from human

plasma as described by Krarup et al. (31) with modifications ofthe affinity chromatography step, as follows. The 4–8% PEGpelletwas dissolved in 145mMNaCl, 5mMCaCl2, 1.5mMNaN3,10 mM Tris-HCl, 0.01% (v/v) emulfogen (Sigma), pH 7.4, andloaded onto an N-acetylcysteine-Sepharose column equili-brated in the same buffer. After washing with (i) the loadingbuffer, (ii) the loading buffer containing 1 M NaCl and 10 mMEDTA instead of CaCl2, and (iii) the loading buffer containing20mMNaCl and 10mMEDTA, bound L-ficolin was elutedwith200mMNaCl, 0.3 MN-acetylglucosamine, 50mMTris-HCl, pH7.8. The eluted proteins were dialyzed against 50 mM NaCl, 10mM CaCl2, 20 mM Tris-HCl, pH 7.8, before further purificationby anion-exchange chromatography.Site-directed Mutagenesis—The expression plasmids coding

for all MASP-3 mutants were generated using theQuikChangeTM XL site-directed mutagenesis kit (Stratagene,La Jolla, CA) according to the manufacturer’s protocol. ThepFastBac1/MASP-3 expression plasmid (19) was used as a tem-plate. Mutagenic oligonucleotides were purchased fromMWG-Biotec (Courtaboeuf, France). The sequences of allmutants were checked by double-stranded DNA sequencing(Genome Express, Grenoble, France).Mass Spectrometry—Analyses were performed using the

matrix-assisted laser desorption ionization technique on aVoy-ager Elite XL instrument (PerSeptive Biosystems, Cambridge,MA) under conditions described previously (32).Crystallization and Data Collection—TheMASP-1/3 CUB1-

EGF-CUB2 segment was concentrated to 3.6 mg/ml in 100 mMnondetergent sulfobetain 195, 145 mM NaCl, 1 mM CaCl2, 50mM triethanolamine-HCl, pH 7.4. Crystals suitable for x-raydiffraction data collection were obtained at 20 °C by the hang-ing drop vapor diffusion method by mixing equal volumes ofthe protein solution and of a reservoir solution composed of18–20% (w/v) PEG 8000, 3% glycerol, and 0.1 M Hepes, pH 7.0.They were transferred to a cryoprotecting solution containing22% (w/v) PEG8000, 20% glycerol, and 0.1 MHepes, pH 7.0, andthen flash-cooled in liquid nitrogen. A native data set indexedin the space group C2221 was measured at the ESRF beamlineID14-eh1 to a resolution of 2.30 Å. The images were processedand the reflections scaled using the XDS program (33). Crystal-lographic statistics for the native data set are given in Table 1.Structure Determination and Refinement—The structure of

CUB1-EGF-CUB2was determined using themolecular replace-ment method. The rotational and translational searches werecarried out using the program AMoRe (34). Although initialsearches using the human Map19 CUB1-EGF structure (20)were unsuccessful, a clear solution was obtained using the ratMASP-2 CUB1-EGF-CUB2 structure in its dimeric form (Pro-tein Data Bank accession code 1NT0) (22). Rigid body refine-ment with the program CNS (35) was used to further improvethe orientation and position of each domain of the searchmodel. Several replacements corresponding to the MASP1/3sequence were clearly seen in the electron density map andintroduced in the model using the graphics program O (36).However, the quality of themap remained unsatisfactory in one

Structure of Interaction Domain of Human MASP-1/3

25716 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 37 • SEPTEMBER 12, 2008


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

of the CUB1 modules, because of a positioning error of its�-barrel, each strand being shifted onto the next one. The cor-rect module orientation was obtained using strong noncrystal-lographic restraints in the model refinement carried out withCNS (35). This essential improvement allowed 489 residues(93% of the model) to be built automatically into the electrondensitymap after two rounds of the ARP-Warp procedure (37),providing clear indication of the quality of the map obtained atthis stage. A new model, free from any bias arising from theinitial molecular replacement model, was thus obtained. Addi-tional model rebuilding and refinement were then carried out.The final stages of the refinement were performed with Ref-mac5 (38), using TLS refinement. The quality of the mapallowed construction of 530 residues out of a total of 554 in theasymmetric unit. Proline residues at positions 17, 183, and 238are in the cis conformation. The atomic coordinates have beendeposited in the Protein Data Bank under the code 3DEM.Modeling of the Interaction between MASP-1/3 CUB1-EGF-

CUB2 and MBL—The homotrimeric collagen-like segment ofhuman MBL containing the putative MASP interaction site(25) was modeled on the basis of published statistical informa-tion derived from collagen-like structures, as described previ-ously (20). FourMBL collagen-like segmentswere dockedman-ually onto the MASP-1/3 CUB1-EGF-CUB2 domain using theinteractive graphics program O (36), with at least one of thethree Lys55 residues of each segment pointing toward a CUB1-EGF-CUB2-binding site. The fact that the MBL collagen-likefibers converge to the same point on their N-terminal side wasused as an additional constraint. The x-ray structure of thezymogen CCP1-CCP2-SP segment of humanMASP-2 (39) wasused as a template for constructing the schematic MBL/MASPmodel depicted in Fig. 7.Surface Plasmon Resonance Spectroscopy—Analyses were

performed using a BIAcore 3000 instrument (GE Healthcare).MBL and ficolins L and Hwere immobilized on the surface of aCM5 sensor chip (GE Healthcare) using the amine couplingchemistry as described previously (18). Binding of the MASP-3variants was measured over 19,300–23,000 resonance units ofimmobilized L-ficolin, 19,600–21,900 resonance units of H-fi-colin, and 7,800–8,000 resonance units ofMBL, at a flow rate of20 �l/min in 145 mM NaCl, 1 mM CaCl2, 50 mM triethanol-amine-HCl, pH 7.4, containing 0.005% surfactant P20 (GEHealthcare). Binding to MBL was measured in the presence of10 mM mannose to prevent unwanted interaction between theoligomannose-type N-linked oligosaccharides of MASP-3 andthe lectin domain of MBL (25). Equivalent volumes of eachMASP-3 sample were injected in parallel over a surface withimmobilized bovine serum albumin to serve as blank sensor-grams for subtraction of the bulk refractive index background.Regeneration of the surfaces was achieved by injection of 10 �lof 1 M NaCl, 20 mM EDTA.

Data were analyzed by global fitting to a 1:1 Langmuirbinding model of both the association and dissociationphases for several concentrations simultaneously, using theBIAevaluation 3.1 software (GE Healthcare). The apparentequilibrium dissociation constants (KD) were calculatedfrom the ratio of the dissociation and association rate con-

stants (koff/kon). Each MASP-3 variant was analyzed at sixdifferent concentrations, ranging from 10 to 250 nM.

RESULTS

Overall Structure—The N-terminal CUB1-EGF-CUB2 inter-action domain of humanMASP-1/3 (amino acids 1–278 of themature proteins) was produced in a baculovirus/insect cellexpression systemas described previously (18).Mass spectrom-etry analysis of the recombinant protein yielded a value of34,238 � 27 Da, accounting for the unmodified polypeptidechain (calculated value, 31,973Da) plus the twoN-linked oligo-saccharides at positions 30 and 159 (total deduced mass,2,265 � 27 Da). The x-ray structure was solved by molecularreplacement using the ratMASP-2 CUB1-EGF-CUB2 structure(22) as a search model, and refined to 2.30-Å resolution. Thefinal Rwork and Rfree factors are 0.22 and 0.24, respectively, andthe refinedmodel has good stereochemistry (Table 1). As antic-ipated from previous ultracentrifugation analyses (17), theCUB1-EGF-CUB2 segment associates as a Ca2�-dependenthomodimer (Fig. 1). The monomers have a curved shape, withthe CUB1, EGF, and CUB2 modules arranged end-to-end. Eachmonomer contains three Ca2� ions, one at each CUB1-EGFinterface (site I) and one bound to one end of the CUB1 andCUB2 modules (sites II and III, respectively). The dimer has ahead-to-tail structure, involving interactions between theCUB1 module of one monomer and the EGF module of itscounterpart (Fig. 1A), i.e. a configuration corresponding to thecompact dimer observed by Feinberg et al. (22). The dimerdisplays noncrystallographic pseudo 2-fold symmetry and hasapproximate dimensions 135 Å (distance between the CUB2modules tips) � 87 Å (distance between the CUB1 modulestips) � 25 Å (mean thickness). On a side view (Fig. 1B), theoverall shape of the dimer is rather flat, the six modules as wellas the Ca2� ions bound to sites II and III being roughly locatedin the same plane.In contrast with human MAp19 (20) which has a well struc-

tured N-terminal extremity, the N-terminal end of the protein(residues 1–6) is disordered in both monomers. In the EGFmodule, most of loop 10 (residues 127–131) is also disordered.Of the two N-linked oligosaccharide chains, the proximal Glc-

TABLE 1Data collection and refinement statisticsData were collected at ID14-eh1, ESRF, Grenoble, France.

Space group C2221Unit cell lengths a � 88.69 Å, b � 110.23 Å, c � 190.59 Å� 0.934 ÅResolution 17 Å to 2.30 ÅRsym 0.08 (0.367)a% completeness 96.2 (90.7)aRedundancy 5.3 (3.7)aI/�(I) average 14.4 (3.5)aNo. of unique reflections 40,219 (6015)a

Model statisticsFinal resolution 2.3 ÅNo. of residues 532No. of water molecules 418No. of ions 6r.m.s.d. �2 bonds 0.005 År.m.s.d. �2 angles 0.87°Rwork 0.22Rfree 0.24

a Statistics for the high resolution bin (2.3–2.44 Å) are given in parentheses.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

NAc residue bound to Asn159 is defined, whereas no density isvisible at Asn30.

The two CUB1-EGF-CUB2 monomers are structurally veryclose to each other, with an r.m.s.d. of 0.5 Å based on 530 C-�atoms. Comparison with the human MAp19 and rat MASP-2structures (20, 22) yields respective r.m.s.d. values of 1.5 Å,based on 292 C-� atoms, and 2.1 Å, based on 496 C-� atoms.These values indicate that there is little variation of module-module orientation at the CUB1-EGF and EGF-CUB2interfaces.The Inter-monomer Interface and Ca2�-binding Site I—The

interface between the two monomers involves a number ofhydrophobic interactions distributed in four major hydropho-bic pockets (Fig. 1A), with a total buried surface of 2100 Å2. (i)Around the 2-fold symmetry axis of the dimer, a central pocketis formed by residues Phe37 and Ile142 frombothmonomers. (ii)At each inter-monomer interface, a second pocket is formed bythe aromatic triad Phe9 (from CUB1), Tyr141, and Tyr146 (fromEGF). Tyr141 also coordinates the Ca2� ion bound to site I,hence providing a link between this site and the inter-monomerinterface. (iii) At both ends of the interface, a distal pocket isformed by residues Tyr17, Pro18, His45 (fromCUB1), and Phe151(from EGF). (iv) A fourth pocket involving residues Tyr42,Met44, and His115 (from CUB1) as well as Tyr137 and His139(from EGF) lies between the distal pocket and the aromatictriad. Although the latter is not present in the C1s CUB1-EGFdimer (40), the four hydrophobic pockets are remarkably con-served in the human MAp19 and rat MASP-2 structures (20,22), and they involve homologous residues (Fig. 2). In MASP-1/3, three additional residues Gln11, Thr110, and Ala118

strengthen the interactions by con-tributing to the intermediary, distal,and central pockets, respectively.Additional interactions between theCUB1 and EGF modules are pro-vided by a direct H-bond betweenHis139 and Tyr42, and by water-me-diated H-bonds between Gln11 andCys147, Asp113 and Ser148, andbetween Ile142 and both Met8and Ala118. As illustrated in Fig. 2,the residues involved in the inter-monomer interface are either con-served or substituted by similar res-idues in the C1r/C1s/MASP family.TheMASP-1/3 EGFmodule has a

topology similar to that describedfor other EGF-like modules (11),with one major and one minor anti-parallel double-stranded �-sheets(Fig. 1A). As observed in ratMASP-2 (22) and human C1r (41),but in contrast with human MAp19(20), loop 10 of MASP-1/3, whichcontains a cluster of charged resi-dues, is mostly disorganized (Figs.1A and 2). The Ca2� ion bound toeach EGF module (site I) is coordi-

nated by seven oxygen ligands, including a water molecule andsix ligands contributed by the EGF module, namely one of theside chain oxygens ofAsp120 andGlu123, the side chain carbonylof Asn140, and the main chain carbonyl of Val121, Tyr141, andGly144. These residues are strictly homologous to thoseinvolved in Ca2� ligation in human C1s (40) and MAp19 (20).Asn140 lacks �-hydroxylation, confirming that insect cells donot achieve this modification (20). As observed in MAp19 andratMASP-2, the watermolecule involved in Ca2� coordinationalso forms an H-bond with Gly36 of module CUB1. AdditionalCUB1-EGF interactions in eachmonomer are achieved by a saltbridge between Arg38 and Glu126, an H-bond between Gly36and Val121, and van derWaals contacts Phe37–Ile142, providingextensive stabilization of the intermodular interface.The CUB1 and CUB2Modules and Ca2�-binding Sites II and

III—As observed previously in the case of human C1s, ratMASP-2, and human MAp19, the CUB modules of humanMASP-1/3 exhibit different folds compared with the canonical�-sandwich topologies containing two five-stranded �-sheetsdescribed initially for plasma spermadhesins (10). Thus, usingthe spermadhesin nomenclature, the CUB2 module lacksstrand�1, whereas CUB1 lacks both�1 and�2. In contrast withthe rat MASP-2 structure (22), CUB2 is equally well defined asCUB1 and does not show a higher temperature factor.A Ca2� ion is bound to the outer end of CUB1 (site II) and to

the inner end of CUB2 (site III) (Fig. 1). In site II, the Ca2� ion iscoordinated by six oxygen ligands, namely both side chain oxy-gens of Asp57, one of the side chain oxygens of Glu49 andAsp102, the main chain carbonyl oxygen of Ser104, and a watermolecule (Fig. 3A). The latter is maintained by H-bonds with

FIGURE 1. Homodimeric structure of the MASP-1/3 CUB1-EGF-CUB2 interaction domain. A, top view.B, bottom view. Monomers A and B are in yellow and red, respectively. Ca2� ions are represented by orangespheres. A, side chains of the major residues involved in the inter-monomer interface are shown.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

the main chain carbonyls of Ser101 and Asp102, providing stabi-lization of the outer end of CUB1. Ca2� coordination in site IIIinvolves six ligands strictly homologous to those listed above

(Asp226, Glu216, Asp263, Ser265, anda water molecule), plus one addi-tional ligand contributed by a sec-ond water molecule (Fig. 3B). Bothwater molecules are stabilized byH-bonds with the main chain car-bonyl oxygen of Ser265, one of themalso interacting with the side chainnitrogens of Asn268 and Asp226, andthe other one forming two H-bondswith Asp263 (Fig. 3).

A comparison of sites II and IIIwith the homologous sites previ-ously described in the CUB1 mod-ules of C1s (40) and MAp19 (20)reveals common features and sev-

eral minor differences. (i) In each case, three acidic residueshomologous to Glu49, Asp57, and Asp102 of MASP-1/3 provideCa2� ligands (Fig. 2). (ii) The residue homologous to Ser104 and

FIGURE 2. Structure-based sequence alignment of selected CUB1-EGF-CUB2 segments. The secondary structure elements and residue numbering arethose of human MASP-1/3 CUB1-EGF-CUB2. Residues involved in the inter-monomer interface are colored yellow. Residues interacting with MBL and ficolinsand involved in the EGF-CUB2 interface are colored blue and pink, respectively. Residues providing Ca2� ligands are in red and marked with closed circles (siteI), open circles (site II), and open squares (site III).

FIGURE 3. Comparative structures of Ca2�-binding sites II (A) and III (B). Oxygen atoms are shown in red,and nitrogen atoms in blue. Water molecules are represented as light blue spheres. Ionic bonds and hydrogenbonds are represented by dotted lines. Loops and �-strands are numbered according to Fig. 2.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Ser265 of MASP-1/3 is a Ca2� ligand in MAp19 but not in C1s.(iii) Asn108 ofMAp19 is aCa2� ligand, unlike its counterparts inother proteins. The latter two differences reflect somemodula-tion of the conformation of loop L9 in these proteins. (iv) Asingle water molecule is involved in the Ca2�-binding site ofMAp19 and in site II of MASP-1/3, whereas two molecules arepresent in C1s and in site III of MASP-1/3. In both sites II andIII, the Ca2� ion is the central element of a network of interac-tions connecting together loops L3 and L9 and strands �5, �6,and �9, hence providing extensive stabilization of the corre-sponding regions of the MASP-1/3 CUB modules. As theircounterparts in C1s and MAp19, Tyr21 of CUB1 and Tyr187 ofCUB2 also play an important part in this network by formingH-bonds with Asp57, Asn105, and Phe109, and with Glu216,Asp226, and Asn268, respectively. Further stabilization of thispart of the CUB modules is achieved by hydrophobic interac-tions involving Phe99, Ser101, and Gly111 (CUB1) and Phe260,Ser262, Arg269, and Gly270 (CUB2).A Structured EGF-CUB2 Interface—Detailed inspection of

the structure at the EGF-CUB2 junction reveals a number ofinteractions between EGF residues mainly located in strands�13 and �14, and CUB2 residues mainly contributed by strand�4� and loop L3� (Fig. 4). These interactions include the follow-

ing : (i) two salt bridges betweenArg163 andGlu165 and betweenAsp135 and Lys189; (ii) direct H-bond between Thr161 andSer190; (iii) four water-mediated H-bonds Tyr153–Pro186,Gly152–Glu165, and between Lys189 and bothAsp135 andCys162;(iv) van der Waals contacts His156–Ser190 and between Tyr153and both Val164 and Pro188. Altogether, these interactionsstrongly stabilize the EGF-CUB2 interface, yielding a buriedsurface of about 750 Å2 at the inter-domain junction. As illus-trated in Fig. 2, many of the residues involved in these interac-tions are conserved in the MASP family, but only a few arepresent in C1r and C1s as well.Mapping of the Interaction Sites with MBL and Ficolins—To

identify the residues of MASP-1/3 involved in the interactionwith MBL and the ficolins, a series of recombinant MASP-3point mutants were produced, and their binding propertieswere analyzed by surface plasmon resonance spectroscopy,using the MASP-3 variants as soluble ligands and immobilizedMBL, L-ficolin, or H-ficolin. Residues of modules CUB1 andCUB2 structurally homologous to those previously found toparticipate in Ca2�-binding site II of human MAp19 (Glu49,Asp102, Ser104, Glu216, Asp226, Asp263, and Ser265) or in its inter-action with MBL and L-ficolin (Tyr56, Glu80, Phe103, Glu106,Tyr225, Glu243, Asn264, and Glu267) (20) were initially targeted

and mutated to alanine. ResiduesHis218 and Glu220, located in a sol-vent-exposed loop in the vicinity ofCa2�-binding site III, were also sub-jected to mutagenesis. As listed inTable 2 and illustrated in Fig. 5,most of these mutations signifi-cantly decreased the ability ofMASP-3 to associate with MBL andficolins L and H. Thus, in moduleCUB1, mutation D102A virtuallyabolished interaction with eachficolin and strongly decreased inter-action with MBL. Mutations E49A,F103A, Y56A, E80A, S104A, andE106A, all decreased interaction

FIGURE 4. Structure of the EGF-CUB2 interface (stereo view). The secondary structure elements and residuesof EGF and CUB2 are shown in green and yellow, respectively. Salt bridges and hydrogen bonds are representedby dotted lines.

TABLE 2Kinetic and dissociation constants for binding of MASP-3 variants to immobilized MBL, L-ficolin, and H-ficolin

MASP-3variant

MBL L-ficolin H-ficolinkon koff KD KD/KD,wt �2 kon koff KD KD/KD,wt �2 kon koff KD KD/KD,wt �2

M�1 s�1 s�1 nM M�1 s�1 s�1 nM M�1 s�1 s�1 nMWild type 6.15 � 105 4.37 � 10�4 0.71 1 2.3 1.82 � 105 4.18 � 10�4 2.30 1 0.8 3.14 � 105 4.68 � 10�4 1.49 1 2.5E49A 1.65 � 105 3.98 � 10�3 24.2 34.1 1.8 1.07 � 105 2.18 � 10�3 20.4 8.9 1.1 9.86 � 104 2.64 � 10�3 26.8 18.0 0.7E49Q 1.81 � 105 3.63 � 10�3 20 28.2 1.8 1.37 � 105 2.78 � 10�3 20.2 8.8 0.4 6.74 � 104 2.17 � 10�3 32.2 21.6 0.6Y56A 1.39 � 105 4.08 � 10�4 2.94 4.1 3.4 7.92 � 104 1.02 � 10�3 12.9 5.6 0.3 1.15 � 105 1.75 � 10�3 15.2 10.2 3.0E80A 2.47 � 105 6.03 � 10�4 2.44 3.4 3.2 8.24 � 104 1.31 � 10�3 15.9 6.9 0.2 1.96 � 105 4.93 � 10�4 2.51 1.7 3.6D102A 4.52 � 104 8.44 � 10�4 18.7 26.3 1.9 NDa ND ND ND ND NDD102N 1.91 � 105 4.41 � 10�3 23.1 27.7 1.0 1.36 � 105 2.63 � 10�3 19.3 8.4 0.3 7.3 � 104 2.65 � 10�3 36.3 24.3 0.4F103A 1.17 � 105 1.10 � 10�3 9.37 13.2 3.4 6.23 � 104 2.34 � 10�3 37.6 16.3 0.3 1.48 � 105 2.23 � 10�3 15.1 10.1 2.8S104A 2.60 � 105 8.16 � 10�4 3.13 4.4 3.8 1.73 � 105 9.31 � 10�4 5.37 2.3 0.2 1.96 � 105 1.40 � 10�3 7.16 4.8 3.1E106A 2.25 � 105 6.88 � 10�4 3.06 4.3 3.6 1.39 � 105 1.11 � 10�3 7.98 3.5 0.2 2.19 � 105 1.08 � 10�3 4.95 3.3 4.0H218A 1.71 � 105 1.72 � 10�3 10.1 14.2 1.4 8.40 � 104 1.48 � 10�3 17.6 7.6 0.6 2.36 � 104 1.83 � 10�3 77.8 52.2 3.4E220A 3.40 � 105 6.18 � 10�4 1.82 2.6 3.1 1.58 � 105 6.07 � 10�4 3.84 1.7 1.5 3.36 � 105 9.65 � 10�4 2.87 1.9 1.9Y225A 1.30 � 105 1.12 � 10�3 8.61 12.1 3.6 4.50 � 104 8.06 � 10�4 17.9 7.8 0.8 ND ND NDE243A 3.18 � 105 4.52 � 10�4 1.42 2.0 4.0 1.71 � 105 1.59 � 10�3 9.27 4.0 1.0 2.58 � 105 5.12 � 10�4 1.98 1.3 4.0N264A 2.69 � 105 2.97 � 10�4 1.11 1.6 2.8 1.30 � 105 1.74 � 10�3 13.4 5.8 0.3 1.72 � 105 6.58 � 10�4 3.82 2.6 2.6S265A 4.37 � 105 4.37 � 10�4 1.0 1.4 2.8 1.29 � 105 2.19 � 10�3 17 7.4 0.3 1.28 � 105 8.69 � 10�4 6.79 4.6 2.3E267A 2.00 � 105 4.29 � 10�4 2.15 3.0 4.0 1.29 � 105 1.97 � 10�3 15.3 6.6 0.3 1.65 � 105 4.71 � 10�4 2.85 1.9 2.7

a Value was not measurable because of the weakness of the binding.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

with either protein, although to varying extents. Interestingly,mutation of Ca2� ligands Glu49 and Asp102 to Gln and Asn,respectively, also strongly decreased their binding properties,resulting in KD values similar to those determined for mutantsE49A and D102A (Table 2). In many cases, the increases in KDmainly resulted from a decrease in the kon value, but the strong-est effects involved both a decrease in kon and an increase in koff.

In module CUB2, mutation to alanine of Asp214 and of thethree acidic residues (Glu216, Asp226, and Asp263) involved inCa2�-binding site III resulted in aggregation of the correspond-ing MASP-3 variants, hence precluding analysis of their inter-action properties. Mutation of the fourth Ca2�-binding ligand,Ser265, had no such effect and decreased interaction with theficolins without significantly altering binding to MBL. Muta-

tion Y225A abolished interactionwith H-ficolin and strongly inhib-ited interaction with MBL and L-fi-colin, whereas H218A markedlydecreased binding to all three pro-teins. Much less pronounced inhib-itory effects were observed formutations E220A, E243A, N264A,and E267A. A second round ofmutations was performed to targetacidic residues found to be exposedin the CUB1-EGF-CUB2 structureand located either in the CUB1(Asp24, Glu55, and Glu107), EGF(Glu126, Glu128, Asp129, Glu130, andGlu131), or CUB2 module (Asp217).None of these mutations had a sig-nificant impact on the interaction ofMASP-3 with either MBL or theficolins, with KD ratios relative towild-typeMASP-3 ranging from 0.7to 1.4 (data not shown).As illustrated in Fig. 6, the muta-

tions that significantly decrease orabolish interaction of MASP-3 withMBL and the ficolins pinpoint resi-dues that are clustered in the vicin-ity of Ca2�-binding sites II and III,thus providing strong support forthe implication of both areas of theCUB1-EGF-CUB2 domain in thebinding.

DISCUSSION

This study provides a secondexample of a CUB1-EGF-CUB2domain structure, allowing a com-parison with that reported previ-ously for rat MASP-2 by Feinberget al. (22). First, it should be empha-sized that our structure is equiva-lent to the compact conformationobserved by these authors, confirm-ing that, of the two distinct dimers

present in their crystal lattice, this one does correspond to thephysiological configuration. In line with the structures of theCUB1-EGF moieties of human MAp19 and C1s (20, 40), ourstudy also provides further evidence of the occurrence of aCa2�-binding site in the CUB modules of the C1r/C1s/MASPfamily, thereby indicating that such sites are indeed present inrat MASP-2 but have been overlooked in the x-ray structurereported by Feinberg et al. (22), as discussed previously in detail(40). A further lesson from the C1s (40), MAp19 (20), ratMASP-2 (22), andMASP-1/3 structures is that all of these pro-teins associate as head-to-tail homodimers through theirCUB1-EGF moieties, by means of an extended inter-monomerinterface. This interface is mainly stabilized by hydrophobicinteractions involving residues that are highly conserved in the

FIGURE 5. Analysis by surface plasmon resonance spectroscopy of the interaction between selectedMASP-3 variants and immobilized MBL, L-ficolin, and H-ficolin. A, interaction with MBL. B, interaction withL-ficolin. C, interaction with H-ficolin. MBL, L-ficolin, and H-ficolin were immobilized on the sensor chip asdescribed under “Experimental Procedures.” The wild-type form and selected variants of the CUB1-EGF-CUB2domain were injected at a concentration of 50 nM. RU, response units.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

C1r/C1s/MASP family and are distributed in four contiguouspockets. Interestingly, the C1s CUB1-EGF dimer differs tosome extent from the other known structures in that it lacksone of these pockets. This may explain why, although purifiedC1s forms homodimers in the test tube, it preferentially associ-ates with C1r under physiological conditions (42, 43), this latterinteraction being possibly more stable than the former.As discussed above, the Ca2�-binding sites harbored by the

CUB modules of C1s, MAp19, and MASP-1/3 exhibit subtledifferences in terms of the number and nature of their coordi-nating ligands, indicating that, as seen for other Ca2�-bindingsites (44), they can slightly adapt their coordinationmode. Nev-ertheless, they have in common a basic framework of interac-tions contributed by a triad of acidic residues (Glu49, Asp57, andAsp102, and Glu216, Asp226, and Asp263 in sites II and III ofMASP-1/3, respectively). With the exception of module CUB2

of C1r where Asp, instead of theexpected Glu, is found at position226, both CUB modules of the pro-teins of the C1r/C1s/MASP familyfeature the consensus Glu-Asp-Aspsequence and may therefore beanticipated to harbor a Ca2�-bind-ing site. Sequence alignment of theknown CUB modules reveals thatmost of thempossess this same con-sensus sequence, indicating that,unlike suggested by initial structuralwork on spermadhesins (10), themajority of the CUB module popu-lation may have the ability to bindCa2�.Twenty seven MASP-3 point

mutants were produced in thisstudy, and analysis of their interac-tion properties provides clear evi-dence that the protease interactswith MBL, L-ficolin, and H-ficolinvia two binding sites contributed bymodules CUB1 and CUB2, andlocated in close vicinity of theirrespective Ca2�-binding sites II andIII (Fig. 6). This conclusion is in linewith our previous study performedon human MAp19, indicating thatthis protein, which only contains aCUB1-EGF module pair, similarlyassociates with MBL and L-ficolinthrough a binding site located inCUB1, in the vicinity of Ca2�-bind-ing site II (20). These findings arefully consistent with our currentknowledge of the interaction ofMASPs with MBL and the ficolins,indicating that this involves in allcases a primary binding site inCUB1and an additional binding site con-tributed by CUB2, the latter being

required to fully stabilize the interaction (18, 21, 45).Mutation of two acidic ligands (Asp60 and Asp105) of Ca2�-

binding site II ofMAp19was previously shown to abolish inter-action with MBL and L-ficolin, and this was interpreted as anindirect effect because of disruption of the Ca2�-binding site,resulting in destabilization of the neighboring interaction site(20). Likewise, this study shows thatmutation of either Glu49 orAsp102, two site II ligands, abolishes or strongly decreases theinteraction properties of MASP-3. Although the implication ofthe corresponding Ca2� ligands in site III could not be testedbecause of the aggregated state of the mutants, this raises thequestion of a direct involvement of some of the Ca2�-bindingresidues in the interaction of the MASPs with MBL and theficolins. This hypothesis appears quite plausible in light of thefollowing observations. (i) Mutagenesis studies performed onhuman and rat MBL (24, 25) and ficolins (26, 27) provide evi-

FIGURE 6. The MBL-ficolin interaction sites of human MASP-1/3. A, top view of the MASP-1/3 CUB1-EGF-CUB2 domain. B, interaction of MASP-1/3 CUB1-EGF-CUB2 with the collagen-like triple helices of tetramericMBL. In each triple helix, one of the three Lys55 residues is shown in black. C and D, detailed views of the regionscomprising Ca2�-binding sites II and III and the proposed interaction sites of MASP-1/3. Mutations resulting instrong inhibition, slight inhibition, and no inhibition are colored red, yellow, and green, respectively. ResiduesGlu128, Asp129, Glu130, and Glu131 are not shown because they belong to a disordered stretch.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

dence that binding of these proteins to theMASPs involves inall cases a major ionic interaction through a conserved lysineresidue (Lys55 in human MBL), implying interaction with anacidic component on the MASP side. (ii) Mutations of Glu80,Glu106, and Glu267 only have a restricted effect on the inter-actions, with KD ratios relative to wild-type MASP-3 rangingfrom 1.7 to 6.9 (Table 2). Most of the other acidic residuesexposed to the solvent in the CUB1-EGF-CUB2 structurehave been subjected tomutagenesis, and none of these muta-tions was effective. A direct implication of acidic ligands ofsite II and/or III in the interactions appears therefore as alogical hypothesis. (iii) Comparison of the four Ca2�-bind-ing CUBmodule structures currently available shows that, ofthe members of the acidic triad, the residues equivalent toGlu49 and Asp102 of the MASP-1/3 site II always coordinateCa2� through one of their carboxyl oxygens (Fig. 3). The freeoxygen groups of Glu49 and Asp102 in CUB1 and of theircounterparts Glu216 and Asp263 in CUB2 point toward theoutside of the domain, roughly to the same direction (Fig. 3,A and B), and appear therefore as ideal candidates for aninteraction with the critical lysine residue conserved in MBLand the ficolins. If this hypothesis is correct, then these res-idues would coordinate Ca2� on one side, and mediate inter-action with MBL or ficolins on the other side, hence provid-ing a major link between the MASPs and their partnerproteins. Interestingly, with the exception of D102A, whichabolishes binding to L- and H-ficolins, all mutations at resi-dues Glu49 or Asp102 inhibit strongly, and to a similar extent,interaction with the three proteins (Table 2), providing sup-port to the hypothesis that both residues contribute equallyto this interaction. Nevertheless, mutations at residues suchas Tyr56, Phe103, His218, or Tyr225 each have a significantinhibitory effect (Table 2), and therefore it appears likelythat these amino acids, all located in close vicinity of site II orIII, also contribute to the interaction to some extent. Fig. 6,Cand D, provides a detailed view of the proposed MBL-ficolininteraction sites of MASP-1/3 featuring the side chains of allresidues thought to participate in the interaction. Given thesimilarities between the present data and those obtainedpreviously onMAp19 (20), and the fact thatmost of the effectivemutations performed in this study had very similar impacts on theinteraction with MBL, L-ficolin, and H-ficolin, it may be antici-pated that the three MASPs associate with each of these proteinsaccording to a common interaction scheme. However, mutation

D105G inMASP-2 abolishes bindingto MBL (45), whereas the corre-sponding mutation D102A inMASP-3 only strongly inhibits inter-action. This suggests that there aresubtle differences between theMASP-2 and MASP-1/3 lineages.Mutagenesis studies on the MBL/fi-colin side lead to the same conclusion(25, 27).The above structural informa-

tion together with previous dataon the nature and location of theMBL interaction site (25) have

been used to build a three-dimensional model of the inter-action between the CUB1-EGF-CUB2 domain of MASP-1/3and the tetrameric form of human MBL (Fig. 6B). As the fourbinding sites defined in the CUB1-EGF-CUB2 homodimer lieapproximately in the same plane, this may associate with MBLin only two alternative orientations, with its flattest side facingeither the C-terminal lectin domains of MBL or its N-terminalextremity. Both orientations have been tested, and the latter(Fig. 6B) was found to better account for the mutagenesis data,particularly for the strong inhibitory effect of the F103A,H218A, and Y225A mutations. An additional advantage of thisconfiguration is that both C-terminal ends of the CUB1-EGF-CUB2 dimer are orientated toward the lectin domains of MBL,allowing the following CCP1-CCP2-SP catalytic domain to foldback into the MBL�MASP complex (see Fig. 7). The resultingCUB1-EGF-CUB2/MBL assembly is symmetrical, each of thefour CUB modules interacting with a collagen-like triple helixof MBL, through a site involving residue Lys55, located abouthalfway along the individual triple helices (Fig. 6B). The dis-tances between the binding sites on the CUB1-EGF-CUB2dimer are roughly similar, with about 55–60 Å between twoadjacent sites, and 75–80 Å between two opposite CUB1 orCUB2 sites. For interaction with the trimeric form of MBL,one of the binding sites of the CUB1-EGF-CUB2 dimer isexpected to be free. Although the interaction symmetry willbe lost, tight binding can still be achieved through the otherthree sites.This type of interaction betweenMBL and theMASP-3 bind-

ing domain has direct implications on the assembly and func-tioning of a whole MBL�MASP complex and leads us to pro-pose a schematic model in which the MASP dimer would beable to sway between a “close” conformation allowing acti-vation of each SP domain by its counterpart (Fig. 7A), and an“open” conformation allowing free access of the SP domainsto their protein substrates (Fig. 7B). Obviously, such a largeconformational change requires a greater extent of flexibilityat the CUB2-CCP1 junction. Although no precise structuralinformation is available in this respect, this hypothesisappears consistent with the observation that this particulararea of MASP-1/3 is highly susceptible to cleavage by pro-teolytic enzymes.4

4 F. Teillet and N. Thielens, unpublished data.

FIGURE 7. Schematic model of a MBL�MASP complex. A, closed conformation. B, open conformation.Domains CUB, EGF, CCP, and SP are colored green, orange, yellow, and purple, respectively. MBL is shownin blue.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Acknowledgments—We thank B. Dublet for mass spectrometry anal-yses, and the scientists at the ESRF beamline ID14-eh1 for their helpwith the use of x-ray diffraction equipment.

REFERENCES1. Holmskov, U., Thiel, S., and Jensenius, J. C. (2003) Annu. Rev. Immunol.

21, 547–5782. Endo, Y., Liu, Y., and Fujita, T. (2006) Adv. Exp. Med. Biol. 586, 265–2793. Ikeda, K., Sannoh, T., Kawasaki, N., Kawasaki, T., and Yamashina, I. (1987)

J. Biol. Chem. 262, 7451–74544. Matsushita, M., and Fujita, T. (1992) J. Exp. Med. 176, 1497–15025. Matsushita, M., Endo, Y., and Fujita, T. (2000) J. Immunol. 164,

2281–22846. Matsushita,M., Kuraya,M., Hamasaki, N., Tsujimura,M., Shiraki, H., and

Fujita, T. (2002) J. Immunol. 168, 3502–35067. Liu, Y., Endo, Y., Iwaki, D., Nakata,M.,Matsushita,M.,Wada, I., Inoue, K.,

Munakata, M., and Fujita, T. (2005) J. Immunol. 175, 3150–31568. Thiel, S., Vorup-Jensen, T., Stover, C. M., Schwaeble, W., Laursen, S. B.,

Poulsen, K., Willis, A. C., Eggleton, P., Hansen, S., Holmskov, U., Reid,K. B., and Jensenius, J. C. (1997) Nature 386, 506–510

9. Dahl, M. R., Thiel, S., Matsushita, M., Fujita, T.,Willis, A. C., Christensen,T., Vorup-Jensen, T., and Jensenius, J. C. (2001) Immunity 15, 127–135

10. Romero,A., Romao,M. J., Varela, P. F., Kolln, I., Dias, J.M., Carvalho, A. L.,Sanz, L., Topfer-Petersen, E., and Calvete, J. J. (1997) J. Mol. Biol. 274,650–660

11. Campbell, I. D., and Bork, P. (1993) Curr. Opin. Struct. Biol. 3, 385–39212. Arlaud, G. J., Barlow, P. N., Gaboriaud, C., Gros, P., and Narayana, S. V.

(2007)Mol. Immunol. 44, 3809–382213. Stover, C. M., Thiel, S., Thelen, M., Lynch, N. J., Vorup-Jensen, T., Jense-

nius, J. C., and Schwaeble, W. J. (1999) J. Immunol. 162, 3481–349014. Takahashi, M., Endo, Y., Fujita, T., and Matsushita, M. (1999) Int. Immu-

nol. 11, 859–86315. Rossi, V., Cseh, S., Bally, I., Thielens, N. M., Jensenius, J. C., and Arlaud,

G. J. (2001) J. Biol. Chem. 276, 40880–4088716. Takahashi, M., Iwaki, D., Kanno, K., Ishida, Y., Xiong, J., Matsushita, M.,

Endo, Y., Miura, S., Ishii, N., Sugamura, K., and Fujita, T. (2008) J. Immu-nol. 180, 6132–6138

17. Thielens, N. M., Cseh, S., Thiel, S., Vorup-Jensen, T., Rossi, V., Jensenius,J. C., and Arlaud, G. J. (2001) J. Immunol. 166, 5068–5077

18. Cseh, S., Vera, L., Matsushita, M., Fujita, T., Arlaud, G. J., and Thielens,N. M. (2002) J. Immunol. 169, 5735–5743

19. Zundel, S., Cseh, S., Lacroix, M., Dahl, M. R., Matsushita, M., Andrieu,J.-P., Schwaeble, W. J., Jensenius, J. C., Fujita, T., Arlaud, G. J., andThielens, N. M. (2004) J. Immunol. 172, 4342–4350

20. Gregory, L. A., Thielens, N.M.,Matsushita,M., Sorensen, R., Arlaud, G. J.,Fontecilla-Camps, J. C., and Gaboriaud, C. (2004) J. Biol. Chem. 279,29391–29397

21. Chen, C. B., and Wallis, R. (2001) J. Biol. Chem. 276, 25894–2590222. Feinberg, H., Uitdehaag, J. C., Davies, J. M., Wallis, R., Drickamer, K., and

Weiss, W. I. (2003) EMBO J. 22, 2348–235923. Wallis, R., and Dodd, R. B. (2000) J. Biol. Chem. 275, 30962–3096924. Wallis, R., Shaw, J. M., Uitdehaag, J., Chen, C. B., Torgersen, D., and

Drickamer, K. (2004) J. Biol. Chem. 279, 14065–1407325. Teillet, F., Lacroix,M., Thiel, S.,Weilguny, D., Agger, T., Arlaud, G. J., and

Thielens, N. M. (2007) J. Immunol. 178, 5710–571626. Girija, U. V., Dodds, A. W., Roscher, S., Reid, K. B., and Wallis, R. (2007)

J. Immunol. 179, 455–46227. Lacroix, M., Arlaud, G. J., and Thielens, N. M. (2007) Mol. Immunol. 44,

393028. Gill, S. C., and von Hippel, P. H. (1989) Anal. Biochem. 182, 319–32629. Teillet, F., Dublet, B., Andrieu, J.-P., Gaboriaud, C., Arlaud, G. J., and

Thielens, N. M. (2005) J. Immunol. 174, 2870–287730. Matsushita, M., Endo, Y., Taira, S., Sato, Y., Fujita, T., Ichikawa, N.,

Nakata, M., and Mizuochi, T. (1996) J. Biol. Chem. 271, 2448–245431. Krarup, A., Thiel, S., Hansen, A., Fujita, T., and Jensenius, J. C. (2004)

J. Biol. Chem. 279, 47513–4751932. Lacroix, M., Rossi, V., Gaboriaud, C., Chevallier, S., Jaquinod, M.,

Thielens, N. M., Gagnon, J., and Arlaud, G. J. (1997) Biochemistry 36,6270–6282

33. Kabsch, W. (1993) J. Appl. Crystallogr. 26, 795–80034. Navaza, J. (2001) Acta Crystallogr. Sect. D 57, 1367–137235. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P.,

Grosse-Kunstleve, R.W., Jiang, J. S., Kuszewski, J., Nilges,M., Pannu,N. S.,Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L. (1998) ActaCrystallogr. Sect. D 54, 905–921

36. Jones, T. A., Zou, J.-Y., Cowan, S. W., and Kjeldgaard, M. (1991) ActaCrystallogr. Sect. A 47, 110–119

37. Morris, R. J., Perrakis, A., and Lamzin, V. S. (2003)Methods Enzymol. 374,229–244

38. Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997)Acta Crystallogr.Sect. D 53, 240–255

39. Gal, P., Harmat, V., Kocsis, A., Bian, T., Barna, L., Ambrus, G., Vegh, B.,Balczer, J., Sim, R. B., Naray-Szabo, G., and Zavodszky, P. (2005) J. Biol.Chem. 280, 33435–33444

40. Gregory, L. A., Thielens, N. M., Arlaud, G. J., Fontecilla-Camps, J. C., andGaboriaud, C. (2003) J. Biol. Chem. 278, 32157–32164

41. Bersch, B., Hernandez, J.-F., Marion, D., and Arlaud, G. J. (1998) Biochem-istry 37, 1204–1214

42. Thielens, N. M., Aude, C. A., Lacroix, M. B., Gagnon, J., and Arlaud, G. J.(1990) J. Biol. Chem. 265, 14469–14475

43. Busby, T. F., and Ingham, K. C. (1990) Biochemistry 29, 4613–461844. Harding, M. (1999) Acta Crystallogr. Sect. D Biol. Crystallogr. 55,

1432–144345. Stengaard-Petersen, K., Thiel, S., Gadjeva, M., Moller-Kristensen, M.,

Sorensen, R., Jensen, L. T., Sjoholm, A. G., Fugger, L., and Jensenius, J. C.(2003) N. Engl. J. Med. 349, 554–560




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

and Nicole M. ThielensFlorence Teillet, Christine Gaboriaud, Monique Lacroix, Lydie Martin, Gérard J. Arlaud

Identification of Its Interaction Sites with Mannan-binding Lectin and Ficolins Domain of Human MASP-1/3 and2-EGF-CUB1Crystal Structure of the CUB

doi: 10.1074/jbc.M803551200 originally published online July 2, 20082008, 283:25715-25724.J. Biol. Chem.

10.1074/jbc.M803551200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/283/37/25715.full.html#ref-list-1

This article cites 45 references, 24 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M803551200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;283/37/25715&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/283/37/25715

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=283/37/25715&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/283/37/25715

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/283/37/25715.full.html#ref-list-1

http://www.jbc.org/

crystal structure of the cub1 -egf-cub2 domain of human masp

Documents