Dynamic Structural Changes During Complement C3 Activation Analyzed by Hydrogen/Deuterium Exchange Mass Spectrometry Michael C. Schustera,1, Daniel Ricklinb,1, Kriszti��n Pappb,2, Kathleen S. Molnarc, Stephen J. Coalesc, Yoshitomo Hamuroc, Georgia Sfyroerab, Hui Chenb, Michael S. Wintersb,3, and John D. Lambrisb,* aDepartment of Medicine, Division of Rheumatology, University of Pennsylvania, Philadelphia, USA bDepartment of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, USA cExSAR Corp., 11 Deer Park Drive, Suite 103, Monmouth Junction, USA Abstract Proteolytic cleavage of component C3 to C3b is a central step in the activation of complement. Whereas C3 is largely biologically inactive, C3b is directly involved in various complement activities. While the recently described crystal structures of C3 and C3b provide a molecular basis of complement activation, they do not reflect the dynamic changes that occur in solution. In addition, the available C3b structures diverge in some important aspects. Here we have utilized hydrogen/ deuterium exchange coupled with mass spectrometry (HDX-MS) to investigate relative changes in the solution-phase structures of C3 and C3b. By combining two forms of mass spectrometry we could maximize the primary sequence coverage of C3b and demonstrate the feasibility of this method for large plasma proteins. While the majority of the 82 peptides that could be followed over time showed only minor alterations in HDX, we observed clear changes in solvent accessibility for 16 peptides, primarily in the ��-chain (�����NT, MG6-8, CUB, TED, C345C domains). Most of these peptides could be directly linked to the structural transitions visible in the crystal structures and revealed additional information about the probability of the structural variants of C3b. In addition, a discontinuous cluster of seven peptides in the MG3, MG6, LNK and �����NT domains showed a decreased accessibility after activation to C3b. Although no gross conformational changes are detected in the crystal structure, this area may reflect a structurally flexible region in solution that contributes to C3 activation and function. Keywords Complement Activation C3 C3b Hydrogen-deuterium exchange Mass Spectrometry *Corresponding author: Tel. +1 215 746 5765, fax +1 215 573 8738, E-mail address: Lambris@mail.med.upenn.edu (J.D. Lambris). 1These authors contributed equally to this work 2Current address: Department of Immunology, E��tv��s Lor��nd University, Budapest, Hungary 3Current address: Division of Infectious Diseases, University of Cincinnati College of Medicine, Cincinnati, USA Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH Public Access Author Manuscript Mol Immunol. Author manuscript available in PMC 2009 June 1. Published in final edited form as: Mol Immunol. 2008 June 45(11): 3142���3151. doi:10.1016/j.molimm.2008.03.010. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

1. Introduction As a central component of innate immunity, the human complement system plays a major role in the recognition and elimination of microbial intruders and other pathogenic cells. Recent research revealed even more important functions for this tuned cascade of soluble and membrane proteins, such as bridging to adaptive immune responses and additional cascades (e.g. coagulation system). Activation of complement component C3 is the point of convergence in the initiation of the complement cascade by the lectin, alternative and classical pathways. This proteolytic conversion leads to the production of the biologically active effector protein C3b and the anaphylatoxin C3a from the biologically inactive protein C3. Covalent attachment of C3b on the surface of foreign cells (i.e. opsonization) induces a variety of terminal complement actions from cell lysis and phagocytosis to the stimulation of downstream immune responses (Markiewski and Lambris, 2007). Since complement activation on host cells has devastating effects and may lead to a number of severe diseases, the cascade activity has to be carefully controlled. The ability to enable and restrict molecular interaction within a single protein template is a central aspect of the control strategy. The biological activities of C3b are directly related to the dynamic exposure of binding sites that are necessary for protein-protein interactions (Gros et al., 2007). While native C3 has a very limited amount of physiological binding partners, C3b gains the ability to bind a variety of essential proteins, including C5 properdin factors B, H and I membrane co-factor protein decay accelerating factor and complement receptor 1 (Sahu and Lambris, 2001). It is through these interactions that C3b and its breakdown fragments are capable of propagating the innate immune response and influencing adaptive immunity. It is not surprising, therefore, that increased activation of C3 or impaired regulation of C3b have been attributed to an increasing number of diseases (Ricklin and Lambris, 2007 Thurman, 2006 Volanakis and Frank, 1998), rendering these proteins as potential targets of therapeutic intervention (Ricklin and Lambris, 2007). In addition, the C3-to-C3b transition is also critical for complement evasion of human pathogens (Lambris et al., 2007). For example, the extracellular fibrinogen-binding protein from Staphylococcus aureus has been shown to preferentially bind the native over the activated form of C3, which may be pivotal for its inhibitory activity (Hammel et al., 2007b). Numerous biological and biophysical techniques have been utilized to investigate the structural changes accompanying the increase in function upon activation of C3. Earlier evidence accumulated through neoepitope mapping with antibodies (Alsenz et al., 1990), chemical modification strategies (Isenman et al., 1981), solution scattering (Perkins and Sim, 1986), and electron microscopy (Smith et al., 1984) indicated significant structural rearrangements during this process. These observations have recently been confirmed by the publication of crystal structures for C3 (Janssen et al., 2005) and C3b (Abdul Ajees et al., 2006 Janssen et al., 2006 Wiesmann et al., 2006) as well as detailed electron microscopy studies (Nishida et al., 2006). While these studies offered a first insight into the mechanism by which C3 activation is propagated, they all were taken under static, non-solute conditions. Even more, the available crystal structures for C3b diverge in some important points and are currently matters of scientific debate (Janssen et al., 2007). Given these structural uncertainties, the versatile biological functions of C3b, as well as the dynamic process in which C3 activity is regulated, more detailed information about how this rearrangement takes place in solution is highly sought after. Hydrogen deuterium exchange (HDX1) coupled with mass spectrometry (MS) has evolved into an indispensable tool for characterizing such dynamic structural changes in solution. HDX takes advantage of the ability of amide backbone hydrogen atoms to exchange with water hydrogens in solution (Busenlehner and Armstrong, 2005 Hoofnagle et al., 2003 Wales and Engen, 2006). When D2O is substituted for H2O in the buffer in which the protein is dissolved, solvent deuterium atoms exchange with backbone hydrogen atoms at a rate influenced by the Schuster et al. Page 2 Mol Immunol. Author manuscript available in PMC 2009 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript

local structure of the protein. Amide hydrogens that are buried within the protein or are involved in hydrogen bonds exchange more slowly with solvent deuterium atoms than do more accessible hydrogen atoms at the protein surface (Englander and Kallenbach, 1983). By analyzing the rate and/or magnitude of the deuterium incorporation into the protein backbone, one can make inferences about the relative structure of the protein. To monitor the amount of deuterium incorporation, a physical technique such as nuclear magnetic resonance (NMR) or MS is employed (Englander, 2006). The choice of analytical technique depends upon the size of the protein NMR is best suited for smaller proteins (less than ~30 kDa) while MS allows for the investigation of proteins in excess of 30 kDa. However, low sequence coverage and other experimental parameters render the analysis of larger proteins increasingly difficult for MS as well (Cravello et al., 2003). Characterization of multidomain plasma proteins such as C3 (184 kDa Fig. 1A) therefore demands an especially high level of instrumental and experimental precision. Although HDX-MS does not provide a three-dimensional structure of the protein, it is capable of reporting structural changes when the protein is studied under varying conditions or in different states (Eyles and Kaltashov, 2004 Hamuro et al., 2003 Schuster et al., 2007). Furthermore, the technique has also been utilized to identify interacting surfaces in protein-protein (Melnyk et al., 2006) and protein-ligand interactions (Hamuro et al., 2006). HDX-MS reflects changes in protein structure that occur in solution but might not be evident from crystallographic structures. As a consequence, the combination of HDX-MS with high-resolution crystal structures can provide an unequaled insight into the solution phase dynamics of proteins. In a previous study, we have successfully used HDX-MS to investigate the structural changes in C3 that take place during its hydrolysis to C3(H2O) (Schuster et al., 2007 Winters et al., 2005). Here we have utilized this approach to investigate the relative solution structures of C3 and C3b. Two MS techniques were employed, matrix-assisted laser desorption/ionization time of flight (MALDI) and electrospray ionization with ion trap (ESI), to maximize the sequence coverage for these large proteins. When analyzed in the context of the available crystal structures for C3 and C3b, our data reveal that there is a cluster of four discontinuous peptides within the MG3, MG6, LNK, and �����NT domains that exhibit increased deuterium exchange in C3 when compared to C3b. The same regions only showed minimal differences in the two crystallographic structures, suggesting that this area may exhibit conformational flexibility in solution. These data not only indicate that this region plays an important role in the activation process but also demonstrate the applicability of the technology to large plasma proteins. 2. Materials and Methods 2.1. Protein Preparation C3 was purified from human plasma (University of Pennsylvania Blood Bank) using established methods (Hammer et al., 1981 Katragadda et al., 2006 Sahu et al., 2000), and was precipitated by dialysis against 5 mM MES buffer (pH 6.0) at 4 ��C. C3b was obtained from C3 by limited digestion with trypsin as previously described (Janssen et al., 2006) and was purified by gel filtration over a Superdex 200 size-exclusion column, followed by anion exchange chromatography over a Mono-Q column. Isolated C3b was treated with iodoacetamide to prevent dimerization, purified over a Mono-S column and then concentrated using an Ultrafree-MC centrifugal filter device. Protein purity was carefully controlled by SDS- PAGE to exclude contaminations, and aliquots of the proteins were stored at ���70 ��C. At the time of use, protein samples was thawed on ice, centrifuged at 3000 rpm in a Sorvall Biofuge Fresco (Thermo Scientific, Waltham, MA) at 4 ��C and reconstituted in the appropriate buffer. Schuster et al. Page 3 Mol Immunol. Author manuscript available in PMC 2009 June 1. NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript