Unique Structure of Ic3b Resolved at a Resolution of 24 Å by 3D-Electron Microscopy,” by Martin Alcorlo, Ruben Martínez-Barricarte, Francisco J

Correction

IMMUNOLOGY Correction for “Unique structure of iC3b resolved at a resolution of 24 Å by 3D-electron microscopy,” by Martin Alcorlo, Ruben Martínez-Barricarte, Francisco J. Fernández, César Rodríguez- Gallego, Adam Round, M. Cristina Vega, Claire L. Harris, Santiago Rodríguez de Cordoba, and Oscar Llorca, which appeared in issue 32, August 9, 2011, of Proc Natl Acad Sci USA (108:13236–13240; first published July 25, 2011; 10.1073/ pnas.1106746108). The authors note that, due to a printer’s error, the affiliation for Adam Round should instead appear as European Molecular Biology Laboratory, 38042 Grenoble, France. The corrected author and affiliation lines appear below. The online version has been corrected.

aCentro de Investigaciones Biológicas, Consejo Superior de Investigaciones Cientíﬁcas, 28040 Madrid, Spain; bCentro de Investigación Biomédica en Enfermedades Raras, 28040 Madrid, Spain; cEuropean Molecular Biology Laboratory, 38042 Grenoble, France; and dDepartment of Infection, Immunity, and Biochemistry, Cardiff University School of Medicine, Cardiff CF14 4X, United Kingdom

www.pnas.org/cgi/doi/10.1073/pnas.1112875108 CORRECTION

www.pnas.org PNAS | September 27, 2011 | vol. 108 | no. 39 | 16481 Downloaded by guest on October 3, 2021 Unique structure of iC3b resolved at a resolution of 24 Å by 3D-electron microscopy

Martin Alcorloa, Ruben Martínez-Barricartea,b, Francisco J. Fernándeza, César Rodríguez-Gallegoa, Adam Roundc, M. Cristina Vegaa, Claire L. Harrisd, Santiago Rodríguez de Cordobaa,b,1, and Oscar Llorcaa,1 aCentro de Investigaciones Biológicas, Consejo Superior de Investigaciones Cientíﬁcas, 28040 Madrid, Spain; bCentro de Investigación Biomédica en Enfermedades Raras, 28040 Madrid, Spain; cStructural Biology Group, European Synchrotron Radiation Facility, 38043 Grenoble, France; and dDepartment of Infection, Immunity, and Biochemistry, Cardiff University School of Medicine, Cardiff CF14 4X, United Kingdom

Edited* by Douglas T. Fearon, University of Cambridge School of Clinical Medicine, Cambridge, United Kingdom, and approved June 20, 2011 (received for review April 28, 2011) Activation of C3, deposition of C3b on the target surface, and organ damage and pathology, complement activation is strictly subsequent amplification by formation of a C3-cleaving enzyme controlled by a number of soluble or membrane-associated (C3-convertase; C3bBb) triggers the effector functions of comple- regulatory proteins [factor H (fH), Decay-accelerating Factor ment that result in inflammation and cell lysis. Concurrently, (DAF), Membrane cofactor protein (MCP), and complement surface-bound C3b is proteolyzed to iC3b by factor I and appropri- receptor 1 (CR1)], which dissociate the C3/C5 convertases and ate cofactors. iC3b then interacts with the complement receptors function as cofactors for the factor I (fI)-mediated proteolysis of (CR) of the Ig superfamily, CR2 (CD21), CR3 (CD11b/CD18), and CR4 C3b (1, 2). Interestingly, although fI-mediated proteolysis inac- (CD11c/CD18) on leukocytes, down-modulating inflammation, en- tivates C3b and helps to preserve complement homeostasis and hancing B cell-mediated immunity, and targeting pathogens for to protect self-components, the C3b degradation products iC3b clearance by phagocytosis. Using EM and small-angle X-ray scatter- and C3dg are also active molecules that interact with specialized ing, we now present a medium-resolution structure of iC3b (24 Å). receptors on leukocytes and are instrumental in modulating the iC3b displays a unique conformation with structural features dis- immune responses and targeting pathogens for clearance by tinct from any other C3 fragment. The macroglobulin ring in iC3b phagocytosis. is similar to that in C3b, whereas the TED (thioester-containing Cleavage of C3b by fI takes place first at two closely located IMMUNOLOGY domain) domain and the remnants of the CUB (complement protein sites in the complement protein subcomponents C1r/C1s, urchin subcomponents C1r/C1s, urchin embryonic growth factor and bone embryonic growth factor and bone morphogenetic protein 1 morphogenetic protein 1) domain have moved to locations more (CUB) domain (Arg1,281-Ser1,282 and Arg1,298-Ser1,299) gen- similar to where they were in native C3. A consequence of this erating iC3b and C3f, a small fragment of 17 amino acids. The fH, large conformational change is the disruption of the factor B bind- MCP, and CR1 are all cofactors of fI for these cleavages. The fI ing site, which renders iC3b unable to assemble a C3-convertase. will then also cleave iC3b between residues Arg932 and Glu933 fi This structural model also justi es the decreased interaction be- generating C3c, which is released into solution, and C3dg, which tween iC3b and complement regulators and the recognition of remains bound to the target. This third cleavage is much slower; iC3b by the CR of the Ig superfamily, CR2, CR3, and CR4. These data under physiological conditions, it is only produced when CR1 further illustrate the extraordinary conformational versatility of serves as a cofactor for cleavage of iC3b by fI (1, 7, 8). C3 to accommodate a great diversity of functional activities. CR2 (CD21) binds iC3b and C3dg, enhancing B-cell immunity (9–11). Similarly, CR3 (CD11b/CD18) and CR4 (CD11c/CD18) omplement is a major component of innate immunity with recognize iC3b and trigger phagocytosis. CR3 and CR4 also Ccrucial roles in pathogen and apoptotic cell clearance, im- perform functions in leukocyte trafficking and migration, synapse mune complex handling, and modulation of adaptive immune formation, and costimulation (12, 13). Notably, phagocytosis responses (1, 2). The complement cascade is triggered by three mediated by binding of iC3b to CR3 is accompanied by down- activation pathways, the classic pathway (CP), the lectin pathway regulation of IL-12 and a lack of oxidative burst in macrophages (LP), and the alternative pathway (AP), which converge in the or by a reduction in the expression of costimulatory molecules central and most important step of complement activation: the and impaired maturation of dendritic cells (12). In addition, C3b formation of unstable protease complexes, called C3 convertases and iC3b bind to a recently described CR of the Ig superfamily (C3bBb in the AP and C4b2a in the CP/LP), that cleave C3 to (CRIg), which also contributes to clearance of pathogens and generate the activated fragment, C3b. When C3b is generated, a apoptotic cells. The expression of CRIg is restricted to a subset of reactive thioester is exposed which is attacked by hydroxyl group- tissue resident macrophages (14). Binding of CRIg to C3b also bearing nucleophiles on adjacent surfaces, resulting in covalent inhibits both AP C3- and C5-convertase activity (15). binding of C3b to the surface. Assembly of the AP C3-convertase Clearly, conversion of C3b into iC3b disrupts domains involved involves Mg2+-dependent binding of factor B (fB) to C3b, in complement amplification and generates CR-specific binding forming the labile proenzyme C3bB; factor D (fD) then cleaves sites that mediate the many iC3b-mediated immunological re- fB to yield the active convertase (C3bBb) (1, 3–6). Convertase- generated C3b forms more C3bBb convertase that cleaves ad- ditional C3 molecules and provides exponential amplification Author contributions: S.R.d.C. and O.L. designed research; M.A., R.M.-B., F.J.F., C.R.-G., and C.L.H. performed research; A.R., M.C.V., and O.L. analyzed data; and S.R.d.C. and O.L. to the deposition of C3b molecules on the pathogen surface. wrote the paper. C3b clustered around these C3 convertases creates an AP C5- The authors declare no conflict of interest. convertase (C3bBbC3b) that cleaves C5. Activation of C5 gen- *This Direct Submission article had a prearranged editor. erates C5a, a potent inflammatory mediator, and C5b, which Data deposition: The 3D-EM map of iC3b has been deposited in the 3D-EM database with triggers the formation of the cytolytic membrane attack complex. accession code EMD-1908. The effector functions of complement, inducing inflammation 1To whom correspondence may be addressed. E-mail: [email protected] or ollorca@ and lysis, contribute to control infection and are clearly an ef- cib.csic.es. fi fective rst-line defense against microbial intruders. However, This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. because a disproportionate complement response may lead to 1073/pnas.1106746108/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1106746108 PNAS Early Edition | 1of5 sponses. The structural bases underlying a similar dynamic rear- In total, 29,730 images of iC3b molecules were collected in the rangement have been deciphered for the activation of C3 and electron microscope after applying fresh protein, immediately other complement components (1–3, 6). The 3D structure of the after elution from a gel filtration column, to an EM grid (Fig. S2). iC3b molecule is presently unclear, precluding similar under- These images were subjected to reference-free classification to standing of the structural rearrangement from C3b to iC3b. More group those images derived from similar views of the protein, than 25 y ago, Isenman (16), using spectroscopic techniques, which were aligned and averaged (Fig. 1B). iC3b appeared on the described that generation of iC3b is accompanied by a major EM support film in different orientations, but a typical view conformational change and suggested a reversion of the change exhibited the macroglobulin (MG) ring (Fig. 1D, blue) with a hole seen in the original conversion of native C3 to C3b. In contrast, at the bottom end of the ring and a wide density at the top (Fig. recent EM data have shown the thioester-containing domain 1D, gray). The comparison between these images and views of C3 (TED) domain of iC3b as a randomly oriented region around the and C3b observed in the electron microscope revealed that the C3c fragment (1, 17), which poses interesting questions regarding TED domain in iC3b is located at a position closest to that oc- the structural basis of the many iC3b-mediated immunological cupied by the TED domain in C3 (Fig. 1E). Notably, the 2D responses (1, 18, 19). averages of iC3b suggested a similar location for the TED domain Here, we have determined the 3D structure of iC3b at medium resolution using 3D-EM and small-angle X-ray scattering in all the molecules collected, which is in clear contrast to the (SAXS). Our data reveal a unique conformation distinct from heterogeneity previously described for the location of the TED that of C3 and C3b. domain in iC3b images obtained in the electron microscope (17). Identical results were obtained for the commercial (Fig. 1) and in- Results house preparations of iC3b (Fig. S2). EM Reveals a Single iC3b Conformation. We have used two prepa- Swing of the TED and CUB Domains in C3, C3b, and iC3b. rations of purified human iC3b: a commercial (Calbiochem) A3D fi preparation that was further purified using Superdex 200 size-ex- structure of iC3b was obtained by angular re nement of the EM clusion chromatography (GE Healthcare) and a preparation of images using an ab initio structure determined by the random iC3b made in our laboratory from in-house prepared human C3 conical tilt (RCT) method as an initial template (Fig. S2). The that was incubated with soluble recombinant MCP and fI and averages obtained after angular refinement were identical to the subsequently purified using anion exchange and gel filtration reference-free averages (Fig. 1 B and C), supporting the correctness chromatography (Fig. 1A and Fig. S1). The chromatographic pro- of the refined structure. The potential conformational heteroge- files and SDS/PAGE analysis of both iC3b preparations were neity of iC3b was evaluated using maximum-likelihood methods indistinguishable. (SI Materials and Methods), which revealed that all the images of

Fig. 1. Purification and EM of iC3b. (A) Final preparations of purified C3, C3b, and iC3b were analyzed by SDS/PAGE. Schematic representation of the primary structure of iC3b is also shown. (B) Reference-free 2D averages of iC3b single-molecule images. Several views of iC3b correspond to a distinct orientation of the protein on the support film. (C) Projections of the EM structure of iC3b obtained by the angular refinement method match the reference-free averages, as shown in B.(D) Selected 2D reference-free average shows a typical view of iC3b. A cartoon highlights the location of the MG ring (blue) and a large density at the top (gray). (E)AsinD, but a typical reference-free average of C3 and C3b molecules observed in the electron microscope is shown (6).

2of5 | www.pnas.org/cgi/doi/10.1073/pnas.1106746108 Alcorlo et al. iC3b collected corresponded to a single or to very closely related between the putative location of the TED domain and the C- conformations (Fig. S3). terminal end of the MG7 domain could hold the remains of the The structure of iC3b at a resolution of 24 Å was analyzed by CUB domain (Fig. 2A, red). computationally fitting the available atomic structures of C3 and C3b within the EM density (Fig. 2A). The MG ring was the most Structure of iC3b in Solution Obtained by SAXS. We performed obvious structural feature, which, at this resolution, was shown to SAXS experiments to determine the overall shape of the C3, C3b, be virtually identical to the atomic structure described for C3b and iC3b molecules using homogeneously purified samples (Fig. (Fig. 2A, blue). Interestingly, the MG ring of C3 did not fit the 1A). Model-independent parameters calculated from the SAXS structure of iC3b at the level of MG7 and MG8 (Fig. 2B). The data were coherent with theoretical values calculated from pub- MG7 and MG8 domains rotate after cleavage of C3, exposing lished crystal structures of C3 and C3b (20, 21) and from the iC3b regions that can interact with certain inhibitors, such as CRIg EM structure (this study) (SI Materials and Methods and Table S1). (15). Thus, the MG ring in iC3b is structurally closer to C3b, from The low-resolution SAXS analysis indicated that the three mole- which iC3b derives, than to C3. The C345C domain could be fi fi cules have well-de ned structures in solution and display limited easily identi ed at the top of the MG ring, but its precise orien- conformational flexibility. The size of C3, C3b, and iC3b calcu- tation could not be determined at this resolution level. We lated from SAXS data gave a radius of gyration and longest di- slightly displaced this domain to accommodate it within the EM mension significantly lower than would be expected for extensively density, because it is accepted that the orientation of this domain SI Materials and Methods varies widely in different structures (Fig. 2A, orange) (1). The disordered species ( ). The theoretical TED domain and the remnants of the CUB domain were located scattering curves for known crystal structures were calculated and fi at the top of iC3b, in the vicinity of C345C, because this EM compared with experimental data. The ttings to the experimental density in our iC3b structure was not accounted by the MG ring data were reasonable even when the structures were considered as SI Materials and Methods and C345C domains. The precise positioning of the TED domain one rigid body ( ). Pair distance distribu- and the remains of the CUB domain could only be hypothesized. tion functions calculated from the SAXS data for C3 and C3b and Modeling based on the computational fitting of their atomic from the SAXS and EM structures of iC3b (Fig. 3A), as well as the structures within the EM density suggested that the TED domain, averaged ab initio models constructed for each of the samples (Fig. which, in C3b, has moved away from the C345C domain, returns 3B), provide support for the notion that iC3b displays a confor- to the proximity of C345C in iC3b (Fig. 2A, green). A density mation significantly closer to C3 than to C3b. IMMUNOLOGY

Fig. 2. 3D structure of iC3b obtained by 3D-EM. (A) 3D structure of iC3b shown as a gray transparent density. The atomic structure of the MG ring (blue) and the C345C domains (orange) were obtained from the structure of C3b (PDB ID code 2I07) (21) and fitted within the EM structure. The putative locations of the TED domain and the remains of the CUB domain in the map have been colored in green and red, respectively. The primary sequence of iC3b chains is also shown. (B) 3D structure of iC3b shown as a gray transparent density, where the MG rings of C3 (PDB ID code 2A73; blue) (20) and C3b (PDB ID code 2I07) (21) have been fitted. The structure of the MG ring in C3 does not fit accurately into the iC3b structure, leaving regions of iC3b unoccupied (labeled as *).

Alcorlo et al. PNAS Early Edition | 3of5 Fig. 3. Global shape of iC3b obtained by SAXS. (A) P(r) functions for C3 (blue), C3b (green), and iC3b (orange) computed from experimental SAXS data. The P(r) function for the EM structure of iC3b (shown in gray) was computed from an equivalent bead model. The P(r) functions are normalized to unity at their maxima. The intersection of each function with the abscissa yields the experimental averaged maximal diameter. P(r), pair distance distribution. (B) Character- istic views of the C3, iC3b, and C3b ab initio models, colored as in A.

Discussion rationale behind these differences could be that the TED and Recent studies have provided a detailed structural view of the CUB regions in iC3b have a strong tendency to rearrange into complexity of the activation and regulation of the complement the conformation we describe here, whereas Nishida et al. (17) AP, illustrating how these processes are driven by large structural captured other possible conformational states generated after rearrangements (1–6, 15, 21–23) (Fig. 4). cleavage of C3b. C3 is a large protein (180 kDa), whose structure results from an The first consequence of the conformational rearrangement intricate arrangement of several domains: a core of 8 homologous from C3b to iC3b is the disruption of the fB binding site on C3b, MG domains forming a ring, a TED domain that contains a re- which renders iC3b unable to assemble the AP C3 convertase. active thioester and connects to the MG core through a CUB We illustrate this, showing an almost imperceptible binding of fB SI domain, the C345C domain that participates in the interaction to iC3b by surface plasmon resonance (SPR; Biacore) ( Materials and Methods with fB, and the N-terminal anaphylotoxin domain (1, 2). Acti- and Fig. S4) and the incapacity of fD to SI Materials and Methods vation of C3 triggers the effector functions of complement, which cleave fB in the presence of iC3b ( and results in inflammation and initiates the lytic pathway. Activation Fig. S5). The C345C domain in C3b binds the von Willebrand of C3 is characterized by a huge conformational displacement of factor type-A (VWA) domain in fB, and a subsequent large the TED domain that exposes a reactive thioester group shielded conformational change in fB is stabilized by interactions between the serine-protease domain in fB and the CUB and MG2 in native C3 (15, 21). In addition, this conformational rearrange- – – ment generates binding sites for a number of molecules like the domains in C3b (1 6, 15, 21 23). Analysis of a disease-associated AP convertase component fB (5, 6) and the complement regu- fB mutant (D279G) that shows increased interaction between lators fH, DAF, MCP, and CR1 (1, 23). the VWA and the C345C domains also failed to show binding to C3b and to generate a C3-convertase (24) (SI Materials and Here, we describe a conformational rearrangement that Methods results in the inactivation of C3b, removing the effector functions and Figs. S4 and S5). As a whole, these data indicate that characterize this molecule. Using EM and SAXS, we dem- that the positioning of the remains of the CUB domain in onstrate that cleavage of C3b into iC3b generates a unique iC3b does not allow stabilization of fB binding, precluding its conformation, distinct from that of C3 and C3b. The overall activation. conformation of the MG ring in iC3b is similar to that in C3b, Disruption of fB binding in iC3b concurs with the generation and the TED domain and the remnants of the CUB domain have of binding sites for CR3 and CR4. These CRs are integrins of the β subfamily that contain a VWA domain involved in iC3b rec- moved to positions more similar to, although different from, 2 ognition (18, 25). The C345C domain in iC3b may still be ca- those occupied by these domains in C3 (Fig. 4). These findings pable of recognizing VWA domains in an ion-dependent substantiate very early work based on spectral changes that manner; thus, we propose that binding of CR3 and CR4 to iC3b suggested a large conformational change in iC3b partially is similar to the interaction between fB and C3b and involves the reverting to a conformation similar to C3 (16). It is important to C345C domain in iC3b. In addition, the rearrangement of the remember that the TED domain is locked in position on surfaces TED domain to the proximity of the C345C domain may con- through the thioester. So, in reality, it is the rest of the iC3b tribute to generate the structural requirements for binding to molecule that moves around while the TED stays stationary. This CR3 and CR4. Supporting this hypothesis, C3d has been shown might be important for CR2, CR3, and CR4 recognition of iC3b to bind weakly to CR3 (26), suggesting that the TED domain of because it implies that a completely different face of the mole- fi iC3b contributes to the interaction with CR3. Also, mutational cule is exposed on the cell surface. In contrast to our ndings, studies have shown that the N terminus of the α′-chain, which is previous EM studies of iC3b have shown that the TED domain proximal to the region occupied by the C345C, CUB, and TED was randomly oriented with respect to the MG ring (1, 17). A domains in iC3b, is involved in the interaction with CR3 (27). CRIg interacts with both C3b and iC3b (15). Our model is compatible with the CRIg binding site remaining accessible after C3b is cleaved into iC3b (Fig. S6). Complement regulators fH, MCP, and CR1 are genetically and structurally related, and they are composed of an array of short consensus repeats (SCRs). SCR4 in fH recognizes the TED domain in C3b (23). Based on the structure of MCP (28), we speculate that SCR4 of MCP may similarly bind the TED domain of C3b and, further, that such a mode of recognition may also hold true for CR1. Thus, the relocation of this domain Fig. 4. Molecular architecture of iC3b. The atomic structures of C3 (PDB ID during the transformation of C3b to iC3b would prevent its in- code 2A73) (20) and C3b (PDB ID code 2I07) (21) were filtered to a resolution teraction with fH and MCP (Fig. S6), as shown before by others similar to that of the structure of iC3b obtained by EM. (29, 30). Similarly, we detect no binding between iC3b and MCP

4of5 | www.pnas.org/cgi/doi/10.1073/pnas.1106746108 Alcorlo et al. and fH using SPR (SI Materials and Methods and Fig. S7) and no Materials and Methods cleavage of iC3b by fI in the presence of MCP and fH (SI Further details are provided in SI Materials and Methods. Materials and Methods and Fig. S8). Under physiological conditions, CR1 is the only complement Purification of iC3b. iC3b was either purchased from Merck (Calbiochem) or regulator that has been described to mediate fI-dependent produced from C3 purified from plasma as described previously (31). cleavage of iC3b (29, 30). Interestingly, the interaction between Biosensor Analysis/Binding Affinity Assays Using SPR. CR1 and C3b is also dramatically affected after the rearrange- All analyses were carried ments in iC3b, indicating that the position of the TED domain in out on a Biacore T100 (GE Healthcare). Further details are presented in SI Materials and Methods. iC3b partly interferes with CR1 binding (SI Materials and Methods fi and Figs. S7 and S8). The higher af nity of CR1 for iC3b com- EM and 3D Reconstruction of iC3b. C3b preparations were negatively stained pared with that of fH or MCP may be critical for CR1 to act as a with 2% (mass/vol) uranyl formate and visualized in a JEOL 1230 trans- cofactor of fI. This may allow CR1 to remain bound to iC3b after mission electron microscope. A total of 29,730 images of iC3b were extracted the first cleavage of C3b, maintaining iC3b in a conformation and refined using either an RCT structure or a featureless Gaussian blob as susceptible to further cleavage by fI. an initial template (Fig. S9). The resolution of the structure was estimated as 24 Å (Fig. S2). The atomic structures of the MG ring from C3 (PDB ID code Concluding Remarks 2A73) (20) and C3b (PDB ID code 2I07) (21) were fitted within the EM density, showing cross-correlation coefficients >0.8. Here, we have resolved the structure of iC3b, which shows a unique conformation among C3 fragments. Our data reveal that SAXS. SAXS data were collected at the BioSAXS station (ID14EH3) at the the TED and CUB domains (and the remnants of the CUB European Synchrotron Radiation Facility. Further details are presented in SI domain after proteolysis) present remarkable motility, adopting Materials and Methods. different positions in the C3, C3b, and iC3b molecules. In contrast, the MG ring appears as a fairly rigid scaffold, supporting ACKNOWLEDGMENTS. The authors acknowledge the European Synchrotron the back and forth swinging of these domains as small fragments Radiation Facility, Grenoble, France, for provision of synchrotron radiation facilities at the ID14EH3 beam line. We thank Prof. B. Paul Morgan for useful of C3 are released by proteolytic cleavage (Fig. 4). The confor- discussions and Prof. Susan M. Lea for soluble recombinant MCP. This work mational rearrangement of C3b into iC3b removes previous was funded by Spanish Ministry of Science and Innovation Grant SAF2008- surfaces present in C3b that are crucial for the assembly of the 00451 (to O.L.); Grant SAF2008-00226 (to S.R.d.C.); and Grants PET2008_0101, BIO2009-11184, and BFU2010-22260-C02-02 (to M.C.V.). It was also funded

AP C3 convertase and for recognition of complement regulators, by the Fundación Ramón Areces (O.L. and S.R.d.C.) and by Grant G0701298 IMMUNOLOGY such as fH and MCP. On the other hand, this rearrangement from the UK Medical Research Council (to C.L.H.). O.L. is additionally sup- generates binding surfaces for the interaction with CR2, CR3 ported by the Red Temática de Investigación Cooperativa en Cáncer from and CR4, which mediate important immunological responses, the Instituto de Salud Carlos III (Grant RD06/0020/1001) and the Human Fron- ﬂ tiers Science Program (Grant RGP39/2008). S.R.d.C. is also supported by the enhancing B cell-mediated immunity, down-modulating in am- Ciber de Enfermedades Raras. M.A. is supported by a Sara Borrell contract mation, and targeting pathogens for clearance by phagocytosis. from the Instituto de Salud Carlos III (CD09/00282).

1. Gros P, Milder FJ, Janssen BJ (2008) Complement driven by conformational changes. 18. Chen X, et al. (2010) Requirement of open headpiece conformation for activation of Nat Rev Immunol 8:48–58. leukocyte integrin alphaXbeta2. Proc Natl Acad Sci USA 107:14727–14732. 2. Rodríguez de Córdoba S, Harris CL, Morgan BP, Llorca O (2011) Lessons from func- 19. Ehlers MR (2000) CR3: A general purpose adhesion-recognition receptor essential for tional and structural analyses of disease-associated genetic variants in the comple- innate immunity. Microbes Infect 2:289–294. ment alternative pathway. Biochim Biophys Acta 1812:12–22. 20. Janssen BJ, et al. (2005) Structures of complement component C3 provide insights into 3. Forneris F, et al. (2010) Structures of C3b in complex with factors B and D give insight the function and evolution of immunity. Nature 437:505–511. – into complement convertase formation. Science 330:1816 1820. 21. Janssen BJ, Christodoulidou A, McCarthy A, Lambris JD, Gros P (2006) Structure of C3b 4. Janssen BJ, et al. (2009) Insights into complement convertase formation based on the reveals conformational changes that underlie complement activity. Nature 444: structure of the factor B-cobra venom factor complex. EMBO J 28:2469–2478. 213–216. 5. Rooijakkers SH, et al. (2009) Structural and functional implications of the alternative 22. Torreira E, Tortajada A, Montes T, Rodríguez de Córdoba S, Llorca O (2009) Co- complement pathway C3 convertase stabilized by a staphylococcal inhibitor. Nat existence of closed and open conformations of complement factor B in the alterna- Immunol 10:721–727. tive pathway C3bB(Mg2+) proconvertase. J Immunol 183:7347–7351. 6. Torreira E, Tortajada A, Montes T, Rodríguez de Córdoba S, Llorca O (2009) 3D struc- 23. Wu J, et al. (2009) Structure of complement fragment C3b-factor H and implications ture of the C3bB complex provides insights into the activation and regulation of the – complement alternative pathway convertase. Proc Natl Acad Sci USA 106:882–887. for host protection by complement regulators. Nat Immunol 10:728 733. 7. Furtado PB, et al. (2008) The partly folded back solution structure arrangement of the 24. Goicoechea de Jorge E, et al. (2007) Gain-of-function mutations in complement factor B 30 SCR domains in human complement receptor type 1 (CR1) permits access to its C3b are associated with atypical hemolytic uremic syndrome. Proc Natl Acad Sci USA 104: and C4b ligands. J Mol Biol 375:102–118. 240–245. 8. Smith BO, et al. (2002) Structure of the C3b binding site of CR1 (CD35), the immune 25. Ustinov VA, Plow EF (2005) Identity of the amino acid residues involved in C3bi adherence receptor. Cell 108:769–780. binding to the I-domain supports a mosaic model to explain the broad ligand rep- 9. Weis JJ, Tedder TF, Fearon DT (1984) Identiﬁcation of a 145,000 Mr membrane protein ertoire of integrin alpha M beta 2. Biochemistry 44:4357–4364. as the C3d receptor (CR2) of human B lymphocytes. Proc Natl Acad Sci USA 81:881–885. 26. Gaither TA, Vargas I, Inada S, Frank MM (1987) The complement fragment C3d fa- 10. van den Elsen JM, Isenman DE (2011) A crystal structure of the complex between cilitates phagocytosis by monocytes. Immunology 62:405–411. human complement receptor 2 and its ligand C3d. Science 332:608–611. 27. Taniguchi-Sidle A, Isenman DE (1994) Interactions of human complement component 11. Matsumoto AK, et al. (1991) Intersection of the complement and immune systems: A C3 with factor B and with complement receptors type 1 (CR1, CD35) and type 3 (CR3, signal transduction complex of the B lymphocyte-containing complement receptor CD11b/CD18) involve an acidic sequence at the N-terminus of C3 alpha′-chain. J – type 2 and CD19. J Exp Med 173:55 64. Immunol 153:5285–5302. 12. Ricklin D, Hajishengallis G, Yang K, Lambris JD (2010) Complement: A key system for 28. Persson BD, et al. (2010) Structure of the extracellular portion of CD46 provides in- immune surveillance and homeostasis. Nat Immunol 11:785–797. sights into its interactions with complement proteins and pathogens. PLoS Pathog 13. Ross GD (2000) Regulation of the adhesion versus cytotoxic functions of the Mac-1/ 6:e1001122. CR3/alphaMbeta2-integrin glycoprotein. Crit Rev Immunol 20:197–222. 29. Lambris JD, et al. (1996) Dissection of CR1, factor H, membrane cofactor protein, and 14. Helmy KY, et al. (2006) CRIg: A macrophage complement receptor required for factor B binding and functional sites in the third complement component. J Immunol phagocytosis of circulating pathogens. Cell 124:915–927. – 15. Wiesmann C, et al. (2006) Structure of C3b in complex with CRIg gives insights into 156:4821 4832. regulation of complement activation. Nature 444:217–220. 30. Ross GD, Lambris JD, Cain JA, Newman SL (1982) Generation of three different ﬁ 16. Isenman DE (1983) Conformational changes accompanying proteolytic cleavage of fragments of bound C3 with puri ed factor I or serum. I. Requirements for factor H vs human complement protein C3b by the regulatory enzyme factor I and its cofactor CR1 cofactor activity. J Immunol 129:2051–2060. H. Spectroscopic and enzymological studies. J Biol Chem 258:4238–4244. 31. Martínez-Barricarte R, et al. (2010) Human C3 mutation reveals a mechanism of dense 17. Nishida N, Walz T, Springer TA (2006) Structural transitions of complement compo- deposit disease pathogenesis and provides insights into complement activation and nent C3 and its activation products. Proc Natl Acad Sci USA 103:19737–19742. regulation. J Clin Invest 120:3702–3712.

Alcorlo et al. PNAS Early Edition | 5of5