Inactivation of human factor VIII by activated protein C: evidence that the factor VIII light chain contains the activated protein C binding site

Factor VIII is represented as a series of heterodimers composed of an 83(81) kDa light chain noncovalently bound to a variable size (93 to 210 kDa) heavy chain. Activated protein C inactivates factor VIII causing several cleavages of the factor VIII heavy chain(s). When factor VIII subunits were dissociated and component heavy and light chains isolated, the heavy chains were no longer a substrate for proteolysis by activated protein C. However, when factor VIII heavy chains were recombined with light chain, the reconstituted factor VIII activity was inactivated by activated protein C. The rate of factor VIII inactivation catalyzed by activated protein C was reduced by the presence of free light chain. The extent of this inhibition was dependent upon the concentration of light chain. Control experiments indicated that this protective effect of free light chain was not the result of inhibition of the activated protein C - lipid interaction. Fluorescence analysis demonstrated binding between the factor VIII light chain, chemically modified with eosin maleimide, and activated protein C, modified at its active site by dansyl-Glu-Gly-Arg chloromethyl ketone. Similar to proteolysis of factor VIII by activated protein C, this binding was dependent upon a lipid surface. Based upon the degree of fluorescence quenching, a spatial distance of 26 A was calculated separating the two fluorophores. These results demonstrate direct binding of activated protein C to the factor VIII light chain and suggest that this binding is an obligate step for activated protein C-catalyzed inactivation of factor VIII.     

The binding of activated protein C to factors V and Va

Activated protein C has been derivatized with the active site-directed fluorophore 2-(dimethylamino)-6-naphthalenesulfonylglutamylglycylarginyl chloromethyl ketone (2,6-DEGR-APC). Covalently modified activated protein C has been used to investigate the binding interactions of the protein to factors V and Va in the presence of phospholipid vesicles. The fluorescence polarization of the 6-dimethylaminonaphthalene-2-sulfonyl moiety increased saturably with increasing phospholipid concentrations in the presence or absence of factor V or Va. Differences in the limiting polarization values indicated distinguishable differences in the interactions between 2,6-DEGR-APC and phospholipid in the presence of factor V or Va. The dissociation constant calculated for the 2,6-DEGR-APC/phospholipid interaction (7.3 X 10(-8) M) was not significantly altered by factor V but was decreased to 7 X 10(-9) M in the presence of factor Va. The interaction between 2,6-DEGR-APC and factor V or Va was characterized by a 1:1 stoichiometry. The binding of 2,6-DEGR-APC to factor V or Va in the presence of phospholipid could be reduced in a competitive manner by diisopropylphosphofluoridate-treated activated protein C. An analysis of the displacement curves indicated that the binding of 2,6-DEGR-APC was indistinguishable from the binding of diisopropylphosphofluoridate-treated activated protein C. The interaction between 2,6-DEGR-APC and phospholipid-bound factor Va was further examined using the isolated subunits of factor Va. Fluorescence polarization changes observed with component E of Va (light chain) closely corresponded with the changes observed with factor Va, whereas isolated component D (heavy chain) had little influence on the binding of 2,6-DEGR-APC to phospholipid vesicles. The data presented are consistent with the interpretation that component E of factor Va contains a binding site for activated protein C.     

The light chain of factor VIII comprises a binding site for low density lipoprotein receptor-related protein.

In the present study, the interaction between the endocytic receptor low density lipoprotein receptor-related protein (LRP) and coagulation factor VIII (FVIII) was investigated. Using purified components, FVIII was found to bind to LRP in a reversible and dose-dependent manner (K(d) approximately 60 nM). The interaction appeared to be specific because the LRP antagonist receptor-associated protein readily inhibited binding of FVIII to LRP (IC(50) approximately 1 nM). In addition, a 12-fold molar excess of the physiological carrier of FVIII, i.e. von Willebrand factor (vWF), reduced the binding of FVIII to LRP by over 90%. Cellular degradation of (125)I-labeled FVIII by LRP-expressing cells ( approximately 8 fmol/10(5) cells after a 4.5-h incubation) was reduced by approximately 70% in the presence of receptor-associated protein. LRP-directed antibodies inhibited degradation to a similar extent, indicating that LRP indeed contributes to binding and transport of FVIII to the intracellular degradation pathway. Degradation of FVIII was completely inhibited by vWF. Because vWF binding by FVIII involves its light chain, LRP binding to this subunit was studied. In ligand blotting experiments, binding of FVIII light chain to LRP could be visualized. More detailed analysis revealed that FVIII light chain interacts with LRP with moderate affinity (k(on) approximately 5 x 10(4) M(-1) s(-1); k(off) approximately 2.5 x 10(-3) s(-1); K(d) approximately 50 nM). Furthermore, experiments using recombinant FVIII C2 domain showed that this domain contributes to the interaction with LRP. In contrast, no association of FVIII heavy chain to LRP could be detected under the same experimental conditions. Collectively, our data demonstrate that in vitro LRP is able to bind FVIII at the cell surface and to mediate its transport to the intracellular degradation pathway. FVIII-LRP interaction involves the FVIII light chain, and FVIII-vWF complex formation plays a regulatory role in LRP binding. Our findings may explain the beneficial effect of vWF on the in vivo survival of FVIII.     

Characterization of a thrombomodulin binding site on protein C and its comparison to an activated protein C binding site for factor Va

Activation of the anticoagulant human plasma serine protease zymogen, protein C, by a complex of thrombin and the membrane protein, thrombomodulin, generates activated protein C, a physiologic anti-thrombotic, anti-inflammatory and anti-apoptotic agent. Alanine-scanning site-directed mutagenesis of residues in five surface loops of an extensive basic surface on protein C was used to identify residues that play essential roles in its activation by the thrombin-thrombomodulin complex. Twenty-three residues in the protein C protease domain were mutated to alanine, singly, in pairs or in triple mutation combinations, and mutants were characterized for their effectiveness as substrates of the thrombin-thrombomodulin complex. Three protein C residues, K192, R229, and R230, in two loops, were identified that provided major contributions to interactions with thrombin-thrombomodulin, while six residues, S190, K191, K217, K218, W231, and R312, in four loops, appeared to provide minor contributions. These protein C residues delineated a positively charged area on the molecule's surface that largely overlapped the previously characterized factor Va binding site on activated protein C. Thus, the extensive basic surface of protein C and activated protein C provides distinctly different, though significantly overlapping, binding sites for recognition by thrombin-thrombomodulin and factor Va.     

Intracellular receptors for activated protein kinase C. Identification of a binding site for the enzyme.

Protein kinase C (PKC) isozymes comprise a family of cytosolic enzymes that translocate to different intracellular sites on activation. We have recently characterized at least two intracellular receptor proteins for PKC (termed RACKs for receptors for activated C-kinase) in the Triton-insoluble material of the particulate fraction from neonatal rat heart. Here, we identify a sequence that appears to resemble the PKC binding site on these RACKs. A peptide (peptide I) with the sequence KGDYEKILVALCGGN bound PKC, and binding was markedly increased in the presence of PKC activators. Furthermore, peptide I inhibited PKC binding to RACKs in a dose-dependent manner. These data suggest that these RACKs have a common PKC binding sequence. Since peptide I inhibited PKC binding to RACKs in vitro, it may be a useful tool to inhibit PKC translocation and subsequent function in vivo.     

Factor VIII light chain contains a binding site for factor X that contributes to the catalytic efficiency of factor Xase

Factor (F) VIII functions as a cofactor in FXase, markedly accelerating the rate of FIXa-catalyzed activation of FX. Earlier work identified a FX-binding site having μM affinity within the COOH-terminal region of the FVIIIa A1 subunit. In the present study, surface plasmon resonance (SPR), ELISA-based binding assays, and chemical cross-linking were employed to assess an interaction between FX and the FVIII light chain (A3C1C2 domains). SPR and ELISA-based assays showed that FVIII LC bound to immobilized FX (K(d) = 165 and 370 nM, respectively). Furthermore, active site-modified activated protein C (DEGR-APC) effectively competed with FX in binding FVIII LC (apparent K(i) = 82.7 nM). Western blotting revealed that the APC-catalyzed cleavage rate at Arg(336) was inhibited by FX in a concentration-dependent manner. A synthetic peptide comprising FVIII residues 2007-2016 representing a portion of an APC-binding site blocked the interaction of FX and FVIII LC (apparent K(i) = 152 μM) and directly bound to FX (K(d) = 7.7 μM) as judged by SPR and chemical cross-linking. Ala-scanning mutagenesis of this sequence revealed that the A3C1C2 subunit derived from FVIII variants Thr2012Ala and Phe2014Ala showed 1.5- and 1.8-fold increases in K(d) for FX, whereas this value using the A3C1C2 subunit from a Thr2012Ala/Leu2013Ala/Phe2014Ala triple mutant was increased >4-fold. FXase formed using this LC triple mutant demonstrated an ~4-fold increase in the K(m) for FX. These results identify a relatively high affinity and functional FX site within the FVIIIa A3C1C2 subunit and show a contribution of residues Thr2012 and Phe2014 to this interaction.     

Inactivation of factor VIII by activated protein C and protein S

Factor VIII was inactivated by activated protein C in the presence of calcium and phospholipids. Analysis of the activated protein C-catalyzed cleavage products of factor VIII indicated that inactivation resulted from the cleavage of the heavy chains. The heavy chains appeared to be converted into 93- and 53-kDa peptides. Inactivation of factor VIII that was only composed of the 93-kDa heavy chain and 83-kDa light chain indicated that the 93-kDa polypeptide could be degraded into a 68-kDa peptide that could be subsequently cleaved into 48- and 23-kDa polypeptides. Thus, activated protein C catalyzed a minimum of four cleavages in the heavy chain. Activated protein C did not appear to alter the factor VIII light chain. The addition of protein S accelerated the rate of inactivation and the rate of all of the cleavages. The effect of protein S could be observed on the cleavage of the heavy chains and on secondary cleavages of the smaller products, including the 93-, 68-, and 53-kDa polypeptides. The addition of factor IX to the factor VIII-activated protein C reaction mixture resulted in the inhibition of factor VIII inactivation. The effect of factor IX was dose dependent. Factor VIII was observed to compete with factor Va for activated protein C. The concentration dependence of factor VIII inhibition of factor Va inactivation suggested that factor VIII and factor Va were equivalent substrates for activated protein C.     

Characterization of a factor Xa binding site on factor Va near the Arg-506 activated protein C cleavage site

Prothrombin is proteolytically activated by the prothrombinase complex comprising the serine protease Factor (F) Xa complexed with its cofactor, FVa. Based on inhibition of the prothrombinase complex by synthetic peptides, FVa residues 493-506 were proposed as a FXa binding site. FVa is homologous to FVIIIa, the cofactor for the FIXa protease, in the FX-activating complex, and FVIIIa residues 555-561 (homologous to FVa residues 499-506) are recognized as a FIXa binding sequence. To test the hypothesis that FVa residues 499-505 contribute to FXa binding, we created the FVa loop swap mutant (designated 499-505(VIII) FV) with residues 499-505 replaced by residues 555-561 of FVIIIa, which differ at five of seven positions. Based on kinetic measurements and spectroscopic titrations, this FVa loop swap mutant had significantly reduced affinity for FXa. The fully formed prothrombinase complex containing this FVa mutant had fairly normal kinetic parameters (k(cat) and K(m)) for cleavage of prothrombin at Arg-320. However, small changes in both Arg-320 and Arg-271 cleavage rates result together in a moderate change in the pathway of prothrombin activation. Although residues 499-505 directly precede the Arg-506 cleavage site for activated protein C (APC), the 499-505(VIII) FVa mutant was inactivated entirely normally by APC. These results suggest that this A2 domain sequence of the FVa and FVIIIa cofactors evolved to have different specificity for binding FXa and FIXa while retaining compatibility as substrate for APC. In an updated three-dimensional model for the FVa structure, residues 499-505, along with Arg-506, Arg-306, and other previously suggested FXa binding sequences, delineate a continuous surface on the A2 domain that is strongly implicated as an extended FXa binding surface in the prothrombinase complex.     

Identification of a binding site for blood coagulation factor Xa on the heavy chain of factor Va. Amino acid residues 323-331 of factor V represent an interactive site for activated factor X

We have recently shown that amino acid region 307-348 of factor Va heavy chain (42 amino acids, N42R) is critical for cofactor activity and may contain a binding site for factor Xa and/or prothrombin [(2001) J. Biol. Chem. 276, 18614-18623]. To ascertain the importance of this region for factor Va cofactor activity, we have synthesized eight overlapping peptides (10 amino acid each) spanning amino acid region 307-351 of the heavy chain of factor Va and tested them for inhibition of prothrombinase activity. The peptides were also tested for the inhibition of the binding of factor Va to membrane-bound active site fluorescent labeled Glu-Gly-Arg human factor Xa ([OG488]-EGR-hXa). Factor Va binds specifically to membrane-bound [OG488]-EGR-hXa (10nM) with half-maximum saturation reached at approximately 6 nM. N42R was also found to interact with [OG488]-EGR-hXa with half-maximal saturation observed at approximately 230 nM peptide. N42R was found to inhibit prothrombinase activity with an IC50 of approximately 250 nM. A nonapeptide containing amino acid region 323-331 of factor Va (AP4') was found to be a potent inhibitor of prothrombinase. Kinetic analyses revealed that AP4' is a noncompetitive inhibitor of prothrombinase with respect to prothrombin, with a K(i) of 5.7 microM. Thus, the peptide interferes with the factor Va-factor Xa interaction. Displacement experiments revealed that the nonapeptide inhibits the direct interaction of factor Va with [OG488]-EGR-hXa (IC50 approximately 7.5 microM). The nonapeptide was also found to bind directly to [OG488]-EGR-hXa and to increase the catalytic efficiency of factor Xa toward prothrombin in the absence of factor Va. In contrast, a peptadecapeptide from N42R encompassing amino acid region 337-351 of factor Va (P15H) had no effect on either prothrombinase activity or the ability of the cofactor to interact with [OG488]-EGR-hXa. Our data demonstrate that amino acid sequence 323-331 of factor Va heavy chain contains a binding site for factor Xa.     

Molecular characterization of an extended binding site for coagulation factor Va in the positive exosite of activated protein C

The anticoagulant human plasma serine protease, activated protein C (APC), inhibits blood coagulation by specific inactivation of the coagulation cofactors factor Va (FVa) and factor VIIIa. Site-directed mutagenesis of residues in three surface loops of a positive exosite located on APC was used to identify residues that play a significant role in binding to FVa. Eighteen different residues were mutated to alanine singly, in pairs, or in triple mutation combinations. Mutant APC proteins were purified and characterized for their inactivation of FVa. Three APC residues were identified that provide major contributions to FVa interactions: Lys(193), Arg(229), and Arg(230). In addition, four residues made significant minor contributions to FVa interactions: Lys(191), Lys(192), Asp(214), and Glu(215). All of these residues primarily contribute to APC cleavage at Arg(506) in FVa and play a small role in the interaction of APC with the Arg(306) cleavage site. In conjunction with previously published work, these results define an extensive FVa binding site in the positive exosite of APC that is primarily involved in binding and cleaving at Arg(506) on FVa.     

The binding of Mn2+ to bovine plasma protein C, des(1-41)-light chain protein C, and activated des(1-41)-light chain activated protein C

The binding isotherms of Mn2+ to bovine plasma protein C (PC), des(1-41)-light chain protein C (GDPC), and activated GDPC (GDAPC) have been measured. PC contains 14-16 total Mn2+ binding sites, a value that is reduced to approximately 7-8 in the presence of NaCl. The average Kd of the latter sites is 230 +/- 30 microM. Upon removal of a 41-residue peptide from the amino terminus of the light chain of PC, and, concomitantly, all of the gamma-carboxyglutamic acid residues, the resulting protein, GDPC, possesses a single Mn2+ site of Kd = 120 +/- 20 microM. Activation of GDPC to GDAPC results in a slight lowering of the Kd for the single Mn2+ binding site to 53 +/- 8 microM, a value that is essentially unchanged in the presence of monovalent cations, a competitive inhibitor of the enzyme, or an active site directed affinity label. The Mn2+ on GDAPC is displaced by Ca2+, suggesting that the protein binding site for these two divalent cations is the same. These studies establish that Mn2+ is a suitable spectroscopic probe for the Ca2+ binding site of GDAPC, and that the divalent cation site is separate from the monovalent cation site(s) and the active site of the enzyme.     

Identification, phosphorylation, and dephosphorylation of a second site for myosin light chain kinase on the 20,000-dalton light chain of smooth muscle myosin

At relatively high concentrations of myosin light chain kinase, a second site on the 20,000-dalton light chain of smooth muscle myosin is phosphorylated (Ikebe, M., and Hartshorne, D. J. (1985) J. Biol. Chem. 260, 10027-10031). In this communication the site is identified and kinetics associated with its phosphorylation and dephosphorylation are described. The doubly phosphorylated 20,000-dalton light chain from turkey gizzard myosin was hydrolyzed with alpha-chymotrypsin and the phosphorylated peptide was isolated by reverse phase chromatography. Following amino acid analyses and partial sequence determinations the second site of phosphorylation is shown to be threonine 18. This site is distinct from the threonine residue phosphorylated by protein kinase C. The time courses of phosphorylation of serine 19 and threonine 18 in isolated light chains follow a single exponential indicating a random process, although the phosphorylation rates differ considerably. The values of kcat/Km for serine 19 and threonine 18 for isolated light chains are 550 and 0.2 min-1 microM-1, respectively. With intact myosin, phosphorylation of serine 19 is biphasic; kcat/Km values are 22.5 and 7.5 min-1 microM-1 for the fast and slow phases, respectively. In contrast, phosphorylation of threonine 18 in intact myosin is a random, but markedly slower process, kcat/Km = 0.44 min-1 microM-1. Dephosphorylation of doubly phosphorylated myosin (approximately 4 mol of phosphate/mol of myosin) and isolated light chains (approximately 2 mol of phosphate/mol of light chain) follows a random process and dephosphorylation of the serine 19 and threonine 18 sites occurs at similar rates.     

Identification of the binding site on S100B protein for the actin capping protein CapZ.

The calcium-binding protein S100B binds to several potential target proteins, but there is no detailed information showing the location of the binding site for any target protein on S100B. We have made backbone assignments of the calcium-bound form of S100B and used chemical-shift changes in spectra of 15N-labeled protein to locate the site that binds a peptide corresponding to residues 265-276 from CapZ alpha, the actin capping protein. The largest chemical-shift changes are observed for resonances arising from residues around the C terminus of the C-terminal helix of S100B and residues Val-8 to Asp-12 of the N-terminal helix. These residues are close to but not identical to residues that have been identified by mutational analysis to be important in other S100 protein-protein interactions. They make up a patch across the S100B dimer interface and include some residues that are quite buried in the structure of calcium-free S100B. We believe we may have identified a binding site that could be common to many S100 protein-protein interactions.     

The myosin light chain enhancer and the skeletal actin promoter share a binding site for factors involved in muscle-specific gene expression.

The myosin light chain (MLC) 1/3 enhancer (MLC enhancer), identified at the 3' end of the skeletal MLC1/3 locus, contains a sequence motif that is homologous to a protein-binding site of the skeletal muscle alpha-actin promoter. Gel shift, competition, and footprint assays demonstrated that a CArG motif in the MLC enhancer binds the proteins MAPF1 and MAPF2, previously identified as factors interacting with the muscle regulatory element of the skeletal alpha-actin promoter. Transient transfection assays with constructs containing the chloramphenicol acetyltransferase reporter gene demonstrated that a 115-bp subfragment of the MLC enhancer is able to exert promoter activity when provided with a silent nonmuscle TATA box. A point mutation at the MAPF1/2-binding site interferes with factor binding and abolishes the promoter activity of the 115-bp fragment. The observation that an oligonucleotide encompassing the MAPF1/2 site of the MLC enhancer alone cannot serve as a promoter element suggests that additional factor-binding sites are necessary for this function. The finding that MAPF1 and MAPF2 recognize similar sequence motifs in two muscle genes, simultaneously activated during muscle differentiation, implies that these factors may have a role in coordinating the activation of contractile protein gene expression during myogenesis.     

Identification of a plasmin-interactive site within the A2 domain of the factor VIII heavy chain

Factor VIII is activated and inactivated by plasmin by limited proteolysis. In our one-stage clotting assay, these plasmin-catalyzed reactions were inhibited by the addition of isolated factor VIII A2 subunits and by Glu-Gly-Arg-active-site modified factor IXa. SDS-PAGE analysis showed that an anti-A2 monoclonal antibody, recognizing the factor IXa-interactive site (residues 484-509), blocked the plasmin-catalyzed cleavage at Arg(336) and Arg(372) but not at Arg(740). Surface plasmon resonance-based assays and ELISA demonstrated that the A2 subunit bound to active-site modified anhydro-plasmin with high affinity (K(d): 21 nM). Both an anti-A2 monoclonal antibody and a peptide comprising of A2 residues 479-504 blocked A2 binding by approximately 80% and approximately 55%, respectively. Mutant A2 molecules where the basic residues in A2 were converted to alanine were evaluated for binding of anhydro-plasmin. Among the tested mutants, the R484A A2 mutant possessed approximately 250-fold lower affinity than the wild-type A2. The affinities of K377A, K466A, and R471A mutants were decreased by 10-20-fold. The inhibitory effect of R484A mutant on plasmin-catalyzed inactivation of factor VIIIa was approximately 20% of that of wild-type A2. In addition, the inactivation rate by plasmin of factor VIIIa reconstituted with R484A mutant was approximately 3-fold lower than that with wild-type A2. These findings demonstrate that Arg(484) plays a key role within the A2 plasmin-binding site, responsible for plasmin-catalyzed factor VIII(a) inactivation.     

Clostridium botulinum C2 toxin. Identification of the binding site for chloroquine and related compounds and influence of the binding site on properties of the C2II channel

Clostridium botulinum C2 toxin belongs to the family of binary AB type toxins that are structurally organized into distinct enzyme (A, C2I) and binding (B, C2II) components. The proteolytically activated 60-kDa C2II binding component is essential for C2I transport into target cells. It oligomerizes into heptamers and forms channels in lipid bilayer membranes. The C2II channel is cation-selective and can be blocked by chloroquine and related compounds. Residues 303-330 of C2II contain a conserved pattern of alternating hydrophobic and hydrophilic residues, which has been implicated in the formation of two amphipathic beta-strands involved in membrane insertion and channel formation. In the present study, C2II mutants created by substitution of different negatively charged amino acids by alanine-scanning mutagenesis were analyzed in artificial lipid bilayer membranes. The results suggested that most of the C2II mutants formed SDS-resistant oligomers (heptamers) similar to wild type. The mutated negatively charged amino acids did not influence channel properties with the exception of Glu(399) and Asp(426), which are probably localized in the vestibule near the channel entrance. These mutants show a dramatic decrease in their affinity for binding of chloroquine and its analogues. Similarly, F428A, which represents the Phi-clamp in anthrax protective antigen, was mutated in C2II in several other amino acids. The C2II mutants F428A, F428D, F428Y, and F428W not only showed altered chloroquine binding but also had drastically changed single channel properties. The results suggest that amino acids Glu(399), Asp(426), and Phe(428) have a major impact on the function of C2II as a binding protein for C2I delivery into target cells.     

Identification of the N-terminal peptide binding site of glucose-regulated protein 94

Because the stress protein GRP94 can augment presentation of peptides to T cells, it is important to define how it, as well as all other HSP90 family members, binds peptides. Having previously shown that the N-terminal half of GRP94 can account for the peptide binding activity of the full-length protein, we now locate this binding site by testing predictions of a molecular docking model. The best predicted site was on the opposite face of the beta sheet from the pan-HSP90 radicicol-binding pocket, in close proximity to a deep hydrophobic pocket. The peptide and radicicol-binding sites are distinct, as shown by the ability of a radicicol-refractive mutant to bind peptide. When the fluorophore acrylodan is attached to Cys117 within the hydrophobic pocket, its fluorescence is reduced upon peptide binding, consistent with proximity of the two ligands. Substitution of His125, which contacts the bound peptide, compromises peptide-binding activity. We conclude that peptide binds to the concave face of the beta sheet of the N-terminal domain, where binding is regulated during the action cycle of the chaperone.     

Localization of the human complement component C3 binding site on the IgG heavy chain

The location of the covalent binding site of the third component of complement (C3) on the IgG heavy chain was determined by sequence analysis of peptides generated by cyanogen bromide digestion of C3-IgG adducts. Activation of the alternative pathway by incubation of heat-aggregated human IgG1 with fresh normal human plasma formed covalent adducts of C3b-IgG. CNBr peptides of these adducts were transferred to a polyvinylidene difluoride membrane, and amino-terminal sequences were determined. A 40-kDa dipeptide containing the covalent bond was identified by labeling the free thiol group (generated during activation of the internal thioester of C3b) with iodo[1-14C]acetamide and analyzed by amino acid sequencing. The resulting double sequence suggested an adduct with NH2 termini at residue 938 (pro-C3 numbering) of C3 (75 residues NH2-terminal to the thioester) and residue 84 in the variable region of the IgG heavy chain. These results combined with results from hydroxylamine treatment (splits ester linkage between C3b and IgG) imply that this adduct peptide consists of a 22-kDa C3 fragment and an 18-kDa IgG fragment. Therefore, C3 binds covalently within the region extending from the last 20 residues of the variable region through the first 20 residues of CH2.