Structural requirements of rabbit IgG F(ab')2 fragment for activation of the complement system through the alternative pathway—I. Disulfide bonds

Selective carboxymethylation of disulfide bonds of the F(ab')₂ fragment from rabbit IgG was performed in order to study its possible implication in the activation of the complement system by the alternative pathway.Disulfide bonds were selectively modified by reduction with different amounts of dithioerythritol (DTE)2 followed by alkylation with iodoacetamide.Monovalent fragments of antibody to sheep erythrocytes, obtained after reduction and carboxymethylation, were capable of mediating the lysis of sheep erythrocytes by the alternative pathway. The chemical modification did not alter the conformational integrity of the protein as shown by CD spectra.These monovolent fragments were as efficient as the bivalent fragment when the results were normalized taking into account the loss of antibody avidity observed in the monovalent antibodies with respect to the bivalent ones.The Fab fragment, obtained by papain digestion, exhibited a similar behaviour to the Fab' fragment. This suggests that the C2-oligosaccharide present in the hinge region of the antibody molecule is not involved in the complement activation through the alternative pathway.These data may suggest that the inter-chain disulfide bonds are not involved in maintaining the effector structures required for complement activation through the alternative pathway.     

Chimeric mouse human IgG3 antibodies with an IgG4-like hinge region induce complement-mediated lysis more efficiently than IgG3 with normal hinge

We have altered the amino acid sequence of the hinge and the first constant domain (CH1) of mouse/human chimeric IgG3 antibodies by site-directed mutagenesis, so as to make the sequences identical to those of IgG4. All the mutant antibodies with altered hinge region were more active in complement activation and complement-mediated lysis than native IgG3. The mutations in CH1, however, did not alter the activity. This demonstrates the importance of the hinge region in modulating this effector function. The results show that the primary structure of neither CH1 nor the hinge of IgG4 is responsible for the lack of complement activation shown by this subclass.     

A mutant human IgG molecule with only one C1q binding site can activate complement and induce lysis of target cells

There are potentially two binding sites for C1q on IgG, one on each C(H)2 domain of the gamma heavy chains, close to the lower hinge region. It is not clear whether the presence and involvement of both the C1q binding sites is necessary to induce the activation signal of human IgG. In order to clarify this issue, we made a hybrid mutant IgG1/IgG3 molecule where the IgG1 half of the molecule was made unable to activate complement through the introduction of a P329A mutation. The IgG3 half of the molecule was mutated to harbor a hinge region identical to that of IgG1, and for detection a peptide tag derived from p21ras was introduced into the FG loop of the C(H)1 domain. The hybrid IgG1P329A/IgG3h1-ras molecules were isolated by Protein A affinity chromatography and shown to activate complement and induce complement-mediated lysis at the same levels as wild-type IgG1 and IgG3h1-ras molecules. Thus, one C1q binding site per IgG is sufficient to induce activation. Wild-type human IgG molecules might also normally expose only one C1q binding site as already shown for interaction with FcgammaR, were IgG expose one binding site per molecule.     

Enhancement of Complement Activation and Cytolysis of Human IgG3 by Deletion of Hinge Exons

The capacity to induce complement-mediated cell lysis is greatly enhanced by truncating the hinge of IgG3 through exon deletions. This was shown by establishing five new cell lines which secreted chimeric IgG3 molecules with specificity for the hapten 4-hydroxy-3-nitrophenacetyl (NP) and having 47,45,32,15, and 0 amino acid hinge regions (the wild-type IgG3 has 62 amino acids in the hinge). Efficient complement activation and complement-mediated cell lysis did not depend on a long total hinge or on a long 'upper' hinge (the stretch from the beginning of the hinge to the first inter-heavy chain S-S bond). On the contrary, the mutant having a 15 amino acid hinge element was up to 10 times more efficient in complement lysis than the wild type. Thus the complement-activation potential appeared to be down-regulated in the wild type. On the other hand, the mutant lacking the hinge altogether did not activate complement or induce complement-mediated cytolysis. These findings have to be taken into account when antibodies are designed for human therapy.     

Activation of complement by human serum IgA, secretory IgA and IgA1 fragments

Activation of the complement (C) system by human IgA was studied. Both subclasses of IgA, IgA1 and IgA2, and secretory IgA were shown to activate C, as determined by deposition of C3 on glutaraldehyde-activated microwells coated with IgA. The activation of the C system occurred in the presence of MgEGTA and not in D-deficient serum. In addition to C3, deposition of properdin (P) but not of C4 was detected. These results indicate that C activation, as determined by measuring deposition of C3 and P, occurred by the alternative pathway (AP). The data further show that the major part of the hinge region, which is deleted in IgA2 as compared with IgA1 and which forms the major structural difference between the two subclasses, is not involved in C activation. Reduction and alkylation destroyed the ability of IgA to activate C, as has also been demonstrated for IgG. In order to define the C activating region of the IgA molecule, several fragments of IgA1 were tested. The four-chain molecules F(ab')2 and F(abc)2 were shown to activate the AP. No activation was observed with the two-chain fragments Fab and Fc. The Fc fragment of IgA also did not activate the CP, as does the Fc fragment of IgG. This indicates that activation of the AP of C by IgA is dependent on the presence of the F(ab')2 fragment. In conclusion: human IgA does activate C by the AP. This activation requires an intact F(ab')2 fragment.     

Antibody dependent cell-mediated cytotoxicity induced by chimeric mouse-human IgG subclasses and IgG3 antibodies with altered hinge region.

A matched set of chimeric mouse-human NP-antibodies were studied for the capacity to induce cell mediated cytotoxicity (ADCC) by normal peripheral blood NK/K cells. The target cells were sheep red blood cells (SRBC) sensitized with the haptens NP or NIP. All four IgG subclasses and several IgG3 variants with altered hinge were tested for ADCC activity. The hierarchy of the ADCC capacity among the subclasses was found to be IgG3 greater than IgG1 greater than IgG4 greater than IgG2. The superiority of IgG3 was only revealed at low effector cell:target cell ratio. The ADCC activity was for the most part unaltered by shortening the hinge region of IgG3 from 62 to 15 amino acids. Also, when the hinge region of IgG3 was mutated to become identical to that of IgG4, the ADCC activity was mainly unchanged. However, an IgG3 variant with deletion of all four hinge exons showed a depressed ADCC activity compared to the wild type. The IgG subclass pattern of complement-mediated lysis (CML) and ADCC is different and the capacity to induce CML and ADCC is changed differently by hinge region modification. Thus CML and ADCC have different structural requirements in the Fc region of IgG.     

Structural Difference in the Complement Activation Site of Human IgG1 and IgG3

The C1q binding epicentre on IgG molecules involves residues Asp²⁷⁰, Lys³²², Pro³²⁹ and Pro³³¹ in the C_H2 domain. IgG1 and IgG3 are usually the most efficient of the four human IgG subclasses in activating complement and they both share all these residues. To reveal possible differences in the structural requirement for complement activation, we created a number of NIP (5-iodo-4-hydroxy-3-nitro-phenacetyl) specific IgG1 and IgG3 antibodies with parallel mutations in or near the putative C1q binding site. The mutants were tested simultaneously for antibody induced, antibody-dependent complement-mediated lysis (ADCML) at high and low antigen concentration on the target cells using sera of human, rabbit and guinea pig as complement source. In addition, we tested the antibodies against target cells decorated with the NP hapten, which has 10-fold lower affinity for the antibodies compared to the NIP hapten. We also used ELISA methods to measure complement activation. We observed a clear difference between IgG1 and IgG3 localized to residues Asp²⁷⁰, Leu³³⁴, Leu³³⁵. For all these residues, and especially for Asp²⁷⁰, IgG1 was heavily reduced in complement activation, while IgG3 was only moderated reduced, by alanine substitution. This difference was independent of the long hinge region of IgG3, demonstrated by hinge region truncation of this isotype such that it resembles that of IgG1. This report indicates the presence of structural differences between human IgG1 and IgG3 in the C1q binding site, and points to a specialization of the two isotypes with respect to complement activation.     

Catabolism of human IgG in mice sensitized to various IgG fragments. Similarities to the catabolism of rheumatoid IgG in mice.

Whole body elimination studies of human serum IgG have showm that C57Bl miceare tolerant to this protein at low concentrations. The present study demonstrates that tolerance to this protein may be broken by presensitization of the mouse with the pepsin-derived fragments of human IgG (F(ab)2 and pFc), in marked contrast to the papain-derived fragments (Fab and Fc). Sensitization with F(ab)2 fragments induced a distinctive elimination pattern of the intact protein which was analogousto that observed in non-sensitized mice injected with serum IgG isolated from patients with rheumatoid arthritis. Since, by circular dichroism studies, we have previouslyimplicated a structural anomaly at or near the hinge region of the 'rheumatoid' IgGmolecule, our observations are discussed in relationship to a possible immune aetiologyfor rheumatoid arthritis.     

Activation of the alternative complement pathway of guinea-gip by liposomes incorporated with trinitrophenylated phosphatidylethanolamine

By incorporation of trinitrophenylamino-caproyldipalmitoylphosphatidylethanolamine (TNP-Cap-DPPE) into liposomes composed of an equimolecular mixture of dimyristoylphosphatidylcholine (DMPC) and cholesterol (TNP-Cap-liposomes), liposomes became readily lysed by guinea-pig serum (GPS) in Mg++-EGTA-GVB (gelatin veronal buffered saline containing 2 mM MgCl2 and 10 mM ethyleneglycol-bis (beta-amino-ethyl ether)N-N'-tetraacetate) as well as in GVB++ (gelatin veronal buffered saline containing 0.15 mM CaCl2 and 0.5 mM MgCl2). Since the classical complement pathway (CCP) does not work in Mg++-EGTA-GVB, TNP-Cap-liposome lysis by GPS in Mg++-EGTA-GVB was thought to be mediated by the activation of the alternative complement pathway (ACP). This conclusion was supported by observations that heating of GPS at 50 degrees impaired its lytic activity while C4-deficient GPS was capable of lytic activity, no lysis occurred in EDTA, and there was noted consumption of complement in GPS treated with TNP-Cap-liposomes at 30 degrees. For TNP-Cap-liposome lysis by GPS in Mg++-EGTA-GVB, the epitope density of the TNP hapten was required to be 5% or more of the DMPC. Changing the acyl group of the phosphatidylcholine (PC) significantly influenced the ACP activating capacity of TNP-Cap-liposome. Dipalmitoyl-PC, DMPC and distearoyl-PC facilitated the ACP activating capacity of the TNP-Cap-liposome, while dilauroyl-PC, egg-PC and dioleoyl-PC did not. Furthermore, the length of spacer between TNP and dipalmitoylphosphatidylethanolamine (DPPE) also influenced the ACP activating capacity and maximum activation was noted when the spacer was aminocaproyl. These physicochemical characteristics which increase the ACP activating capacity coincided with those reported to increase the immunogenicity of hapten-sensitized liposomes.     

The antigen binding domain of non-idiotypic human anti-F(ab')2 autoantibodies: study of their interaction with IgG hinge region epitopes.

In previous studies we described a natural human IgG-anti-F(ab')2 autoantibody family with immunoregulatory properties. Genes coding for the variable regions of the heavy and light chains of the Abs were isolated from a natural Ig gene library and scFv Abs were expressed in E. coli. The scFv Abs bound to F(ab')2 but not to Fab fragments. This points to an epitope located in the hinge region since Fab fragments are lacking most of the hinge. In order to verify our hypothesis, double chain peptides comprising the lower-, middle-, and part of the upper hinge subregion of IgG1-IgG4 were synthesized on cellulose membranes and tested for binding to the Abs. The results show binding of Abs to IgG1 and IgG4 hinge region peptides. In order to identify the key residues of the discontinuous epitopes we carried out complete substitutional analyses in which each amino acid of the wt peptides was substituted by all other amino acids except cysteine. The exchange of proline in the IgG1 or IgG4 middle hinge region abrogated the binding, revealing the importance of this subregion for epitope expression. No binding to the IgG2 or IgG3 hinge was detected. These results indicate that scFv anti-F(ab')2 Abs recognize the hinge region of IgG1 and IgG4 and that the expression of the epitope depends on an intact middle hinge subregion.     

Enzymatic Fragmentation of an Unusual Human IgG2 (Kva) Myeloma Protein

The Kva IgG2(k) myeloma protein showed a complete resistance to papain in the presence of cysteine at neutral pH, and a higher resistance to trypsin and alpha-chymotrypsin digestion than other IgG2 proteins. On the other hand, the Kva molecule was cleaved by pepsin at low pH to give the expected F(ab')2 fragment. When the cleavage conditions were altered, it was possible to obtain Fab, Fc, and Fc' fragments from this molecule as well. The Fab/c fragment and FacbFc' complex were also obtained, which have not previously been reported from human IgG2 molecules. Incubation at elevated temperatures (45-50 degrees C) and/or lower pH resulted mainly in enzymatic attack on the C terminal side of the hinge. It was necessary to destroy the hinge by reduction or to expose the Kva molecule at 70 degrees C or at lower pH (2.5) prior to digestion to facilitate enzyme digestion on the NH2 terminal side of the hinge. These results indicate that the hinge region of the Kva molecule has an unusually compact structure, which makes it extremely resistant to proteolysis.     

Role of the antibody Fc in the immune clearance of Trypanosoma cruzi

Passive transfer of immune serum obtained from mice chronically infected with Trypanosoma cruzi to mice containing circulating bloodstream trypomastigotes induces a very fast clearance of the parasites. Comparison of trypomastigotes clearance in normocomplementemic and C5-deficient mice showed no difference. IgG fraction obtained from immune serum was very efficient at inducing complement-mediated lysis and immune clearance of bloodstream trypomastigotes, whereas its Fc-missing F (ab') 2 fragments, although able to induce lysis, were unable to induce clearance. It is suggested that the immune clearance of bloodstream trypomastigotes is dependent on the antibody Fc region and that complement-mediated lysis is not a prerequisite for elimination of the parasites from circulation.     

Conformation of the Hinge Region and Various Fragments of Human IgG3

A study of the circular dichroic (CD) spectra of various fragments of human IgG3, including the isolated hinge region, Fh, has shown that the hinge region has a high degree of an unusual secondary structure, unique within immunoglobulin material recorded to date. This structure appears to be rigid and aperiodic throughout the hinge region and is compatible with a repeated amino acid sequence. The conformation of the isolated Fh fragment is the same as that of the bound hinge region; also, there is no substantial conformational interaction between the hinge region and the Fab or Fc fragments of human Igtg3. a comparison of the CD spectra of Fc and pFc fragments isolated from an IgG1 and an IgG3 myeloma protein has shown that subclass differences of amino acid sequence do not substantially alter the conformation of these fragments.     

Interaction of the hinge region of human immunoglobulin G with a murine lymphocyte membrane receptor. Relevance to the problem of antiglobulin induction in rheumatoid arthritis.

Evidence is presented which indicates the presence on murine lymphocytes of a membrane receptor for determinants on the hinge region of human IgG. These determinants are exposed following pepsin scission of IgG molecule, i.e. on the F(ab')2 fragment. The evidence for a hinge receptor derives, in vivo, from splenic localization of F(ab')2 in germinal centres and, in vitro, from immunofluorescent binding studies. The sequential immunofluorescent pattern for the uptake of human F(ab)2 fragments into murine spleen germinal centres was identical with that previously observed for heat-aggregated human IgG, but F(ab')2 fragments appeared to be retained in the germinal centres for a shorter time than aggregated IgG. Experiments with nude mice and T cell-deprived mice showed that the localization of F(ab')2 fragments does not require T cells. Competition experiments suggest that the receptor for F(ab')2 may bear little relation to the receptor for aggregated IgG. The relevance of such a lymphocyte membrane receptor to the immunopathology of rheumatoid arthritis is discussed in the light of previous findings that a proportion of the serum IgG of patients with rheumatoid arthritis has a structural anomaly compared with control IgG, characterized exposure of new determinants at the hinge region.     

Activation of complement by human IgG1 and human IgG3 antibodies against the human leucocyte antigen CD52.

Activation of the complement cascade by immunoglobulin G (IgG) plays a major role in the host defense against pathogens. Using recombinant human antibodies specific for the leucocyte antigen CD52, different allotypes of human IgG1 subclass were compared for their ability to activate human complement. In addition the roles of the different length hinge regions of IgG1 and IgG3 were investigated. It was found that the naturally occurring allotypes G1m(a,z) and G1m(f), and one artificially created isoallotype, G1m(null), did not significantly differ in their overall ability to cause cell lysis. However, some differences in binding of individual components of the classical activation pathway were detected. More of the complement component C1s seemed to be associated with the allotype G1m(f), although this did not result in an overall improvement in lytic potency. In this system the wild-type IgG3 was found to be less effective in complement lysis than IgG1. By shortening the hinge region of IgG3 to resemble that of an IgG1 antibody, increased complement binding was observed compared with that of wild-type IgG3 and the IgG1 allotypes. The overall lytic potency of the antibody was also improved compared with wild type IgG3 and it was also slightly more effective than the IgG1 allotypes.     

Preparation and biologic characterization of fragments containing dimeric and monomeric C gamma 2 domain of rabbit IgG

A number of fragments derived from acid-treated rabbit IgG by digestion with plasmin have been separated by high-resolution gel filtration. Fragments isolated included a dimer and monomer Facb, named F(acb)2 and Facb, respectively and a heterodimer composed of Facb and Fab subunits, named F(acb)(ab). A C gamma 2 fragment was obtained by papain digestion of Facb. A heterodimer composed of Facb and Fab', named F(acb)(ab'), was also prepared by oxidizing a reduced mixture of these fragments. Fragments thus obtained are classified into two groups--those carrying paired C gamma 2 domains, i.e. F(acb)2, and the disulfide-linked dimeric C gamma 2 fragment; and those having a single C gamma 2 domain, i.e. reduced, alkylated Facb and C gamma 2 fragment, F(acb)(ab) and F(acb)(ab'). These fragments exhibited marked differences in their capacity to activate complement in assay systems of hemolysis and complement consumption by immune complexes or aggregates on polystyrene latex. Fragments of the former group could activate complement but with a definitely reduced efficiency (50%) compared to intact IgG, whereas fragments of the latter group were practically inactive. Although it was not determined whether the C1-binding capacity itself is changed by monomerization of the C gamma 2 domain, the results suggested that the intact paired C gamma 2 module is required at least for the activation process of complement.     

Lung injury induced by antibody fragments to angiotensin-converting enzyme.

Rabbits given goat anti-rabbit angiotensin-converting enzyme antibodies or derived antibody fragments develop rapidly fatal pulmonary edema. Endothelial cell injury is manifested by bleb formation and the disintegration of cell membranes. Platelets are found along the injured endothelium and leukocytes block capillary lumens. The pathologic features are similar when immune IgG, F(ab')2, or Fab are given. In vitro studies of complement activation show that solubilized, purified angiotensin-converting enzyme alone activates C1, with consumption of C4 and C3. Addition of immune IgG plus converting enzyme enhances this activation. F(ab')2 plus enzyme enhances only C3 consumption, while Fab with enzyme produces no additional complement utilization. Thus, while complement activation may be involved in the pathogenesis of injury induced by IgG or F(ab')2, the mechanism of Fab-induced endothelial injury remains unclear.     

Activation of the alternative complement pathway by natural antibody to glycolipids in guinea-pig serum.

Liposomes containing paragloboside (PG) on their membrane were readily lysed by C4-deficient guinea-pig serum (C4D-GPS) through activation of the alternative complement pathway (ACP). Therefore we examined the reactivity of several types of guinea-pig serum (GPS) on PG-liposomes and determined that all GPS except that from specific pathogen-free (SPF) Hartley guinea-pigs had lytic capacity in Mg-EGTA-GVB (gelatin veronal-buffered saline containing Mg++ and ethyleneglycol-bis(beta-aminoethyl ether)N,N''-tetraacetate). This lytic capacity of GPS corresponded with the amount of natural antibody to PG in those sera. Although GPS of SPF guinea-pigs (SPF-GPS) could not lyse PG-liposomes in Mg-EGTA-GVB, it could lyse the liposomes when heated C4D-GPS or Hartley GPS was added. Natural antibody to PG in the heated sera was regarded to have sensitized PG-liposomes to lysis by SPF-GPS via ACP activation. Since the antibody to PG-liposomes was removed by lacto-N-nor-hexaosylceramide which has the same chemical structure in the terminal oligosaccharide, the antibody to PG in GPS was suggested to have a specificity to the terminal structure of oligosaccharide shared by lacto-N-nor-hexaosylceramide. Furthermore, the IgM fraction, which had been prepared by gel filtration of heated C4D-GPS on a Sephadex G200 column, could also sensitize PG-liposomes to lytic reaction of SPF-GPS in Mg-EGTA-GVB. This sensitizing capacity of heated C4D-GPS was suppressed by absorption of the serum or its IgM fraction with anti-guinea-pig mu-chain antibody coupled to Sepharose. Therefore, it was concluded that the lysis of PG-liposomes by GPS in Mg-EGTA-GVB was a result of ACP activation mediated by natural antibodies to PG of the IgM type which are present in usual GPS. This conclusion indicated that natural antibodies of the IgM type might play a role with ACP in host defence, especially in C4-deficient guinea-pigs where the classical complement pathway is impaired.     

The effect of C3 depletion on the clearance of Trypanosoma cruzi induced by IgG antibodies

Passive transfer of homologous immune serum or IgG anti-Trypanosoma cruzi antibodies to normal mice containing circulating T. cruzi bloodstream trypomastigotes (Btrys) induces a very fast clearance of the parasites. Previous work from this laboratory has shown that F(ab')2 fragments obtained from IgG anti-T. cruzi antibodies retain the ability to induce lysis of the Btrys but are unable to induce clearance of the parasites. This suggests that the clearance is dependent on the Fc region. The removal of the Btrys may then be effected by attachment of the antibodies to the parasites and removal of the opsonized parasites by the mononuclear phagocytic system. Attachment of the opsonized parasites to macrophages may be effected either through the Fc receptor that binds specifically to the heavy chain of IgG or through C3b fragments after cleavage of C3 by C3-convertases. In order to find out the possible role of C3b in this phenomenon the clearance of Btrys was determined in mice depleted of C3 by previous treatment with cobra venom factor. The results of these experiments showed that depletion of C3 completely abolished the immune clearance induced by anti-T. cruzi antibodies. It is suggested that C3 is required for the clearance of Btrys from circulation.     

Alteration of the conformation of human IgG subclasses by reduction of the hinge S-S bonds

Human IgG changed molecular size upon mild reduction and alkylation as shown by HPLC gel filtration. IgG1, IgG2 and IgG4 proteins increased in molecular size while IgG3 proteins were decreased in molecular size by this treatment. Several proteins within each subclass covering different light chain types and Gm types were tested all showing the same effect. A plausible explanation was related to the hinge and to the CH2 region since Fab fragments experienced unchanged molecular size irrespective of IgG subclass while Fc (of IgG1, IgG2, IgG3, containing only two aa of the 62 aa long hinge of IgG3 and IgG4) increased in size and Fch (which contains most of the 62 aa long hinge region of IgG3) decreased in size upon reduction and alkylation. It is postulated that reduction of the hinge S-S bonds permit the IgG and Fc molecules to open up in the CH2 region due to the lack of trans-interaction here, resulting in a larger molecular size. For IgG3 and Fch (from IgG3) molecules there was an opposite and even greater effect on the open polyproline like structure of the gamma 3 hinge which depends on intact S-S bonds (there are 11 bonds here). Reduction of these S-S bonds apparently breaks down this open hinge structure resulting in a net decrease in molecular size of IgG3 and Fch molecules.