Molecular studies of rheumatoid factor using pseudobioaffinity membrane chromatography

Using pseudobioaffinity ligand L-histidine immobilized to poly(ethylene vinyl alcohol) hollow fiber membrane is an interesting approach for the purification of total IgG and its subclasses from untreated serum in a single step. Gentle adsorption and elution conditions of this chromatography system allow efficient recovery of the protein in its native form. This approach was employed for the recovery and molecular study of rheumatoid factor (RF), an anti-IgG autoantibody (AAb) that form immune complexes with autologous IgG Abs in the sera. The purity of the recovered molecule was analyzed on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), revealed a 150-kDa IgG band and an additional approximately 300 kDa band which may be RF bound IgG complex. Since RF is an AAb, the purified protein was studied for its catalytic functions like peptide, DNA, and RNA hydrolyzing activities. The substrate Pro-Phe-Arg-4-methyl-coumaryl-7-amide (PFR-MCA) hydrolyzing activity by total IgG from different patient sera was found to be greater than healthy controls. In an effort to identify the subclass specificity for the proteolytic function, the pre-purified total IgG fractions from rheumatoid arthritis (RA) sera were subjected to rechromatography using a discriminating buffer. In this experiment, the activity was found in the non-retained fractions suggesting IgG2 specificity for the catalytic function. A comparative study between different catalytic functions was performed for IgG separated from individual patient.     

Human monoclonal IgG rheumatoid factor has structural homology with bacterial Fc receptor proteins

A cloned lymphoblast cell line, hRF-1, that secreted human monoclonal IgG4 rheumatoid factor autoantibody was produced by Epstein-Barr virus transformation of lymphocytes from rheumatoid arthritis synovium. The binding of hRF-1 rheumatoid factor to IgG globulins of different mammalian species was similar to the binding specificity of Staphylococcus aureus protein A (SpA) and to antibodies found in the sera from patients with rheumatoid arthritis. hRF-1 also had the same binding pattern to human IgG subclasses as SpA. Direct competition was observed between SpA and hRF-1 in binding IgG Fc. These results provide evidence for structural homology between a bacterial Fc receptor protein (SpA) and the monoclonal IgG rheumatoid factor.     

Quantitative evaluation of serum IgG subclass levels by an immunoenzyme method

We developed a enzyme linked immunosorbent assay (ELISA) for measuring IgG subclasses concentration in serum. For this we used monoclonal antibodies. The specificity of these antibodies was evaluated with a panel of myeloma proteins belonging to the 4 IgG subclasses. The ELISA was sensitive (allowing the detection of subclasses at ng level) and accurate (inter-assay coefficient of variation of 14%). Using the WHO serum 67/97 as reference, we determined the concentration of IgG subclasses in a pool of sera. In addition concentrations were measured in 69 healthy adults to study the distribution of each IgG subclass. A good correlation (r = 0.78) was obtained between the sum of the subclasses measured by ELISA and total IgG measured by immunonephelometry.     

A monoclonal anti-idiotype specific for human polyclonal IgM rheumatoid factor.

One of the hallmarks of rheumatoid arthritis (RA) is the production of high titers of rheumatoid factor (RF) antibody directed against the Fc portion of IgG. Anti-Id that recognize the majority of monoclonal RF from patients with B cell dyscrasias are reactive with only 1 to 2% of these polyclonal RF from RA patients. We describe a new monoclonal anti-Id, 4C9, that recognizes a L chain determinant on polyclonal IgM RF from patients with RA but does not recognize a panel of monoclonal RF from patients with B cell malignancies. 4C9 reactivity is found in the serum of 34/43 RF-positive RA patients and in 12/12 RF-positive synovial fluids, but in only 1/14 RF-negative sera from RA patients and 1/22 sera containing monoclonal IgM RF. 4C9 reactivity is highly enriched in purified IgM RF from nine RA patients and represents a variable percentage of total IgM RF up to a maximum of 23%. Furthermore, 4C9 reactivity is enriched in the synovial fluid of three of five RA patients compared with serum, suggesting that 4C9-reactive IgM RF are synthesized within the joint. IgG RF from RA synovial fluids are not 4C9 reactive, indicating either that different genes are used to encode IgM and IgG RF in RA patients, or that IgG RF have somatically mutated away from idiotypic reactivity.     

The amino acid sequences of the three heavy chain constant region domains of a human IgG2 myeloma protein.

The amino acid sequences of most of the CH1, CH2 and CH3 domains of IgG Zie, a myeloma protein belonging to the IgG2 subclass, are presented. These data make possible a comparison of the sequences of residues 253-446 of all four subclasses of immunoglobulins: these residues make up almost the entire Fc regions. A comparison can also be made of the CH1 domain of IgG1 Eu and the CH1 domain of IgG2 Zie. Earlier sequence analyses of the Fc regions of subclass 1 and 3 proteins, and parts of the Fc regions of subclass 2 and 4 proteins showed that about 95% of these sequences were identical. The extended comparisons made possible by the data presented here show that this very high degree of identity is maintained throughout the four subclasses. Similarly, the CH1 domains of gamma 1 and gamma 2 chains were found to have about 93% sequence identity. It is unlikely that the few single amino acid changes within the constant region domains can account for the marked differences between subclasses observed in the region domains can account for the marked differences between subclasses observed in the biological effector functions of immunoglobulin Fc regions, especially since most of the changes are highly conservative. Rather, it seems probable that these functional differences are caused by conformational differences between the subgroups, which result from sequence differences in the hinge regions.     

Rheumatoid factors may bear the internal image of the Fc gamma-binding protein of herpes simplex virus type 1

Affinity-purified rheumatoid factors (RF) from 20 patients with rheumatoid arthritis were tested for their reactivity with the mAb II-481 against glycoprotein E (gE), the Fc gamma-binding protein of HSV-1, as well as with a panel of mAb against human Fc gamma R. All RF bound to mAb II-481 in preference to mAb IV.3 (anti-human Fc gamma RII) or MOPC 141 (control mAb) which belong to the same IgG2b subclass. Five RF showed strong reactivity with II-481. No significant reactivity was observed between RF and mAb against human Fc gamma R. Non-RF human IgM did not react with any of the mAb. Clear-cut binding to II-481 was also seen with monoclonal IgM-RF derived from MRL/1 mice (mRF-2). The reaction between RF and II-481 was completely inhibited by human IgG. It was also inhibited by BHK cell extract infected with HSV-1, and with purified gE. II-481 inhibited the binding of human IgG Fc to the infected cell extract, confirming that II-481 recognizes the Fc-binding site on gE. II-481 did not react directly with human IgG or Fc of IgG. mAb to human IgG2 showed stronger binding to II-481 than to MOPC 141, suggesting II-481 has conformational similarity to human IgG H chain. These results suggest that at least some RF bear the "internal image" of HSV-1 Fc gamma-binding protein and support the hypothesis that some RF may be generated as anti-idiotype antibodies against antiviral antibodies.     

Rheumatoid factors from patients with rheumatoid arthritis react with beta 2-microglobulin.

IgM rheumatoid factors (RF) from 18 sera of rheumatoid arthritis (RA) patients isolated from monomeric IgG affinity columns showed strongly positive ELISA reactions with human beta 2-microglobulin (beta 2m), as well as with recombinant beta 2m. When the same RA sera were adsorbed to beta 2m-Sepharose affinity columns, eluted material showed predominant IgM anti-Fc of IgG and anti-beta 2m reactivity. Inhibition reactions with "RF" obtained from IgG affinity columns showed slightly higher reactivity of RF for Fc over beta 2m; however, when RF from the same RA serum had been adsorbed to and eluted from beta 2m affinity columns, beta 2m showed greater inhibition than Fc for RF reacting with either beta 2m or Fc on ELISA plates. Thus two overlapping populations of RF were identified in RA sera showing reactivity with both beta 2m and Fc of IgG. When RF were isolated from IgG columns, affinity was slightly higher for Fc than beta 2m. Conversely, RF eluted from beta 2m Sepharose reacted slightly more with beta 2 m than Fc. Trypsin digests of a polyclonal RA IgM RF showed no beta 2m reactivity in Fc mu 5 fragments. Fab mu RF retained slight anti-Fc IgG but no residual anti-beta 2m activity. Monoclonal human IgM, IgG, or IgA RF either from mixed cryoglobulins or EBV-stimulated RA lymphoid cell lines showed negative or occasional weakly positive anti-beta 2m activity. Overlapping 7-mer peptide ELISA analysis of the entire 99-amino acid sequence of beta 2m showed a major RF-reactive linear hydrophilic sequence at positions 56-60 which included a 3-amino acid exact homology to positions 401, 403, and 404 of the C gamma 3 domain. A peptide encompassing this sequence produced 90% inhibition of RF binding to whole beta 2m. Substitution of neutral glycines for each amino acid throughout the reactive epitope at positions 56-66 indicated that lysine at position 58 aspartic acid at 59, and tryptophane at 60 represented major portions of the RF-reactive epitope. These findings indicate that human RF derived from patients with RA react with other epitopes besides those present on IgG Fc, including epitopes on human beta 2m. For many years serum RF3 found in patients with RA have been regarded as premier examples of autoantibodies to autologous IgG.(ABSTRACT TRUNCATED AT 400 WORDS)     

Streptococcal Fc receptors. II. Comparison of the reactivity of a receptor from a group C streptococcus with staphylococcal protein A

The reactivity of a soluble Fc receptor from a group C streptococcus ( FcRc ) was compared antigenically and functionally with the staphylococcal Fc receptor, protein A. Protein A and FcRc were found to inhibit each others' binding to the Fc region of human IgG, indicating that they bind to sites that are in close proximity on the Fc region of human IgG. The two bacterial Fc receptors were antigenically unrelated. Differences were observed in the species and subclass reactivity of the two receptors. The patterns of binding of protein A and FcRc under various conditions suggested that these receptors reacted with distinct regions on the Fc region of immunoglobulins. FcRc bound more efficiently to goat, sheep, and cow IgG, protein A bound more efficiently to dog IgG, and neither receptor bound to rat IgG. Differences were also observed in the reactivity towards human IgG subclasses. The FcRc bound to all samples of the four human IgG subclass standards. Protein A bound to IgG1, IgG2, and IgG4, and to one of two IgG3 myeloma proteins tested. The reactivity of our soluble FcRc corresponds to a type III streptococcal Fc receptor classified by the reactivity of intact bacteria.     

Modulation of human rheumatoid factor-specific lymphocyte responses with a cross-reactive anti-idiotype bearing the internal image of antigen

Rabbit anti-idiotypic antibodies to human rheumatoid factor (RF) autoantibodies were isolated by affinity chromatography on rabbit anti-human IgG Fc Sepharose 4B. The anti-idiotypic antibodies bore the "internal image" of the antigen, human IgG. They reacted specifically with multiple human monoclonal and polyclonal IgM-RF, independent of any particular light or heavy chain amino acid sequence. The anti-idiotypes did not react with IgM or IgG proteins lacking RF activity. The present experiments determined the potential of the "internal image" antibodies to modulate in vitro lymphocyte functions. The addition of anti-idiotypic antibody to peripheral blood mononuclear cell cultures from patients with rheumatoid arthritis elicited lymphocyte proliferation, but not RF synthesis. The antibody did not induce the proliferation of lymphocytes from a normal individual. Moreover, the anti-idiotype specifically suppressed IgM-RF secretory responses when preincubated with B cells before co-culture with autologous pokeweed mitogen-activated T cells. The data show that the anti-idiotypic antibodies with the "internal image" of antigen are capable of interacting with B cell receptors in an antigen-restricted manner, and possess specific immunomodulatory properties.     

Mapping rheumatoid factor binding sites using genetically engineered, chimeric IgG antibodies.

We are using chimeric IgG antibodies consisting of murine variable regions joined to human constant regions as rheumatoid factor (RF) binding substrates to localize and map IgM RF binding sites on IgG. Using chimeric antibodies in a modified RF ELISA, we showed that RFs from rheumatoid arthritis (RA) and Waldenstrom's macroglobulinemia (WMac) patients differ in their binding specificities for IgG3, although some of these RFs share common specificity for IgG1, IgG2, and IgG4. By shuffling constant region domains between IgG3 and IgG4, we showed that sequence variation in the CH3 domain is responsible for WMac-derived RF differentiation of IgG3 and IgG4. By making site-directed mutations in the wild-type IgG3 or IgG4 human gamma constant genes, we showed that His-435 is an essential residue in RF binding to IgG for most WMac RFs. The allotypic polymorphism in IgG3 at 436 is not responsible for differences in previous reports of high-frequency IgG3 binding by WMac RFs. A amino acid loop in the CH2 domain of IgG4 proximal to the CH2-CH3 interface is important in WMac RF binding to IgG; a more distal CH2 loop in CH2 has a more variable effect on WMac RF binding. To evaluate the contribution of the N-linked carbohydrate moiety at Asn-297 to RF binding sites on IgG, we measured RF binding to aglycosylated IgG antibodies produced by mutating the glycosylation signal Asn-297 to another amino acid. Of all four IgG subclasses, only aglycosylated IgG3 was a better RF binding substrate than its glycosylated subclass counterpart.(ABSTRACT TRUNCATED AT 250 WORDS)     

Crystal structure of a human autoimmune complex between IgM rheumatoid factor RF61 and IgG1 Fc reveals a novel epitope and evidence for affinity maturation

Rheumatoid factors (RF) are autoantibodies that recognize epitopes in the Fc region of immunoglobulin (Ig) G and that correlate with the clinical severity of rheumatoid arthritis (RA). Here we report the X-ray crystallographic structure, at 3 A resolution, of a complex between the Fc region of human IgG1 and the Fab fragment of a monoclonal IgM RF (RF61), derived from an RA patient and with a relatively high affinity for IgG Fc. In the complex, two Fab fragments bind to each Fc at epitopes close to the C terminus, and each epitope comprises residues from both Cgamma3 domains. A central role in the unusually hydrophilic epitope is played by the side-chain of Arg355, accounting for the subclass specificity of RF61, which recognizes IgG1,-2, and -3 in preference to IgG4, in which the corresponding residue is Gln355. Compared with a previously determined complex of a lower affinity RF (RF-AN) bound to IgG4 Fc, in which only residues at the very edge of the antibody combining site were involved in binding, the epitope bound by RF61 is centered in classic fashion on the axis of the V(H):V(L) beta-barrel. The complementarity determining region-H3 loop plays a key role, forming a pocket in which Arg355 is bound by two salt-bridges. The antibody contacts also involve two somatically mutated V(H) residues, reinforcing the suggestion of a process of antigen-driven maturation and selection for IgG Fc during the generation of this RF autoantibody.     

A rheumatoid factor paradox: inhibition of rituximab effector function

Introduction

Rituximab (RTX) therapy of rheumatoid arthritis (RA) exhibits enhanced effectiveness in seropositive patients. Using patient sera, we tested if this improved efficacy was associated with enhanced RTX mediated complement-dependent cytotoxicity (RTX-CDC).

Methods

We developed an in vitro assay for RTX-CDC using patient sera and the Daudi human B cell line. Using propidium iodide uptake and flow cytometry, we compared RTX-CDC with rheumatoid factor (RF)+ sera relative to normal volunteer, non-RA and RF- sera. Additional studies examined mixing studies of RF+ and RF- sera, as well as the effect of monoclonal IgA or IgM RF. Finally, the effect of RF on RTX mediated trogocytosis of normal B cells was evaluated.

Results

Using human sera, addition of RTX resulted in rapid and profound (> 50%) Daudi cell death that was complement dependent. Surprisingly, RF+ patient sera exhibited reduced RTX-CDC relative to RF- sera, with an inverse relationship of RTX-CDC and RF titer. Mixing studies indicated the presence of an inhibitor of RTX-CDC in RF+ sera. The addition of monoclonal IgM or IgA RF to RF- sera markedly inhibited RTX-CDC. This effect was specific for RF binding to the Fc portion of RTX as it was not apparent with the F(ab)'' domains of RTX engineered onto IgG3 heavy chain. RF also modestly inhibited RTX mediated trogocytosis.

Conclusions

Contrary to expectations, RF+ sera exhibits reduced RTX-CDC due to the presence of RF. The enhanced efficacy of RTX in seropositive RA patients cannot be attributed to improved B cell depletion through CDC. This result indicates that high RF levels may potentially modulate the efficacy of any therapeutic monoclonal antibody dependent on Fc effector function.     

T15 group A streptococcal Fc receptor binds to the same location on IgG as staphylococcal protein A and IgG rheumatoid factors

Previous work has shown that IgG rheumatoid factors (RF) bind to the C gamma 2-C gamma 3 interface region of human IgG in the same area that binds staphylococcal protein A (SPA). Group A, C, and G strains of Streptococci possess Fc receptors that bind to IgG but not to fragments containing only the C gamma 2 or C gamma 3 domains. This work describes the binding site location on human IgG for the binding of the isolated Fc receptor from the T15 strain of a Group A streptococcus and its relationship to the site that binds SPA and the IgG RF. The isolated T15 Fc receptor (T15) with a molecular mass of 29.5 kD inhibited the binding of IgG RF to IgG. The binding of T15 itself to IgG was strongly inhibited by SPA (42.0 kD) and its monovalent fragment D (7 kD). Human IgG fragments consisting of the C gamma 3 domains did not inhibit the binding of T15 to IgG, whereas those with both domains were effective inhibitors. T15 did not bind to rabbit IgG fragments consisting of either the C gamma 2 or C gamma 3 domains, but did bind to those with both domains. An IgG3 myeloma protein was a poor inhibitor and has been shown to bind poorly to the IgG RF. Most IgG3 myeloma proteins did not bind to SPA. The substitution of Arg and Phe for His 435 and Tyr 436 is responsible for the poor binding of IgG3 to SPA and to the IgG RF. Chemical modification of His or Tyr on IgG reduced its ability to inhibit the binding of T15 to IgG. Reversal of the chemical modifications with hydroxylamine resulted in near complete restoration of inhibitory capacity. This information, collectively, coupled with the known positions in space of the His and Tyr residues in the C gamma 2-C gamma 3 interface region, verified that both His 435 and Tyr 436, and possibly His 310 and 433, are involved. These residues are also involved in binding SPA and the IgG RF. These data therefore indicate that the T15 Group A Streptococcal Fc receptor binds to the same location on the Fc of IgG as SPA and the IgG RF. The biologic relevance of these similarities between bacterial cell wall Fc receptors and IgG RF are not yet apparent, but suggest that RF could bear the internal image of these bacterial structures.     

Radioimmunoassay of IgG and IgM rheumatoid factors reacting with human IgG.

Although IgG rheumatoid factor may play a central role in the pathogenesis of rheumatoid arthritis, previously there have been no precise methods for its specific measurement in serum and synovial fluid. This paper describes a solid phase radioimmunoassay for the independent quantification of IgM and IgG rheumatoid factor reacting with the Fc fragment of human IgG. As measured by this assay, serum IgG rheumatoid factor levels differed significantly between patients with seropositive and seronegative rheumatoid arthritis and normal control subjects. In addition, several sera and joint fluids from patients with seropositive rheumatoid arthritis, even without vasculitis, were shown by gel chromatography to have acid-dissociable complexes of IgG rheumatoid factor suggestive of IgG-IgG dimer or trimer formation.     

Antigenic specificities of human monoclonal and polyclonal IgM rheumatoid factors. The C gamma 2-C gamma 3 interface region contains the major determinants

The binding site specificity of 12 monoclonal and 11 polyclonal IgM rheumatoid factors (RF) isolated from human plasma or serum has been studied. All IgM RF bound best to sites on IgG and intact Fc. The monoclonal IgM RF did not bind at all to fragments lacking the C gamma 2 or C gamma 3 domains. In contrast, low level binding to the pFc' fragment, composed of the C gamma 3 domain, was seen with seven IgM RF, mainly from patients with rheumatoid arthritis (RA). IgG1 binding appeared to be a requisite specificity of all human IgM RF. IgM RF binding to IgG3 subclass was common among the monoclonal IgM RF. Most RA polyclonal IgM RF but only 2 of the monoclonal IgM RF possessed the IgG1, 2 and 4 binding pattern. Monoclonal IgM RF which bound best to histidine-modified IgG also bound well to IgG3. The 7-kDa fragment D of staphylococcal protein A inhibited the IgG binding of most monoclonal and to a lesser degree polyclonal IgM RF. Thus, the results indicate that the C gamma 2-C gamma 3 interface region of IgG contains the predominant determinants for monoclonal and polyclonal IgM RF. For some monoclonal IgM RF the binding site, even though at the interface of the C gamma 2 and C gamma 3 domains, is not the staphylococcal protein A site. Furthermore, polyclonal IgM RF possess specificities not encountered among the monoclonal IgM RF. These specificities may have special     

Indices of anti‐dengue immunoglobulin G subclasses in adult Mexican patients with febrile and hemorrhagic dengue in the acute phase

Heterologous secondary infections are at increased risk of developing dengue hemorrhagic fever (DHF) because of antibody‐dependent enhancement (ADE). IgG subclasses can fix and activate complement and bind to Fc? receptors. These factors may also play an important role in the development of ADE and thus in the pathogenesis of DHF. The aim of this study was to analyze the indices of anti‐dengue IgG subclasses in adult patients with febrile and hemorrhagic dengue in the acute phase. In 2013, 129 patients with dengue fever (DF) and 57 with DHF in Veracruz, Mexico were recruited for this study and anti‐dengue IgM and IgG determined by capture ELISA. Anti‐dengue IgG subclasses were detected by indirect ELISA. Anti‐dengue IgG2 and IgG3 subclasses were detected in patients with dengue. IgG1 increased significantly in the sera of patients with both primary and secondary infections and DHF, but was higher in patients with secondary infections. The IgG4 subclass index was significantly higher in the sera of patients with DHF than in that of those with DF, who were in the early and late acute phase of both primary and secondary infection. In conclusion, indices of subclasses IgG1 and IgG4 were higher in patients with DHF.

     

Immunoglobulin G subclasses of anti-thyroid peroxidase autoantibodies in human autoimmune thyroid diseases

A distribution of immunoglobulin G (IgG) subclass of anti-thyroid peroxidase (TPO) autoantibodies was studied to know whether anti-TPO autoantibodies are closely implicated in the pathogenesis of human autoimmune thyroid diseases. As a result of analyzing 14 patients' sera, 7 with Graves' disease and 7 with Hashimoto's thyroiditis, anti-TPO autoantibodies were found to consist of mainly IgG1 subclass. Percentages of both IgG1 and IgG2 subclasses in IgG class of autoantibodies corresponded to those in the normal serum composition, whereas IgG3 subclass was scarcely contained in anti-TPO autoantibodies and IgG4 subclass markedly increased. It was thought that anti-TPO autoantibodies had a capability to lyse thyroid follicular cells by the mechanism of antibody-dependent complement-mediated cytolysis, because IgG1 and IgG2 subclasses of antibodies can fix complement and TPO locates in apical membrane surface of thyroid follicular cells. Comparing Graves' disease with Hashimoto's thyroiditis, mean percentages of both IgG1 and IgG2 subclasses of 2 groups were statistically different. Namely, sera of patients with Graves' disease had higher and lower mean percentages of IgG1 and IgG2 subclasses of autoantibodies, respectively, than those with Hashimoto's thyroiditis, though no plausible explanation for these differences can be offered at the present time.     

Measurement of serum levels of natalizumab, an immunoglobulin G4 therapeutic monoclonal antibody

Human immunoglobulin G4 (IgG4) is a poor trigger of effector functions and, therefore, is the preferred subclass for therapeutic monoclonal antibodies that merely aim to block their in vivo targets. An example is natalizumab, a recombinant IgG4 antibody directed against α4-integrin and used for treatment of multiple sclerosis. Efficient treatment requires that the pharmacokinetics of therapeutic monoclonal antibodies can be accurately monitored. For natalizumab, this requires special precautions due to recently reported structural peculiarities of human IgG4. Here we describe the development of an assay to determine serum levels of natalizumab. Compared with other IgG subclasses, human IgG4 possesses unique structural properties that influence its interactions in both in vivo and in vitro settings. Thus, IgG4 undergoes Fab arm exchange in vivo, resulting in effectively monovalent antibodies. Furthermore, IgG4 is able to bind to other human and nonhuman IgG via Fc interactions. We demonstrate how these features can interfere with measurement of specific IgG4 and describe how we addressed these issues, resulting in an assay that is not sensitive to Fab arm exchange by natalizumab or to IgG4 Fc interactions.     

Fc and Fab Fragments from IgG<Subscript>2</Subscript> Human Immunoglobulins Characterized

Papain digestion of 7S immunoglobulin G (IgG) produces two 3.5S Fab fragments and one 3.5S Fc fragment^1–8. The Fab fragment contains one light chain and one Fd fragment and is still able to combine specifically univalently with antigen. The Fc fragment is a dimer of the carboxyl terminal half of the heavy chain. Pepsin splits 7S IgG into some small peptides derived from Fc and one 5S F(ab′)₂ fragment, which contains both antigen-binding sites. Based on this information, some investigators^6,7 have postulated that pepsin splits the γ chains at the C-terminal side of the inter-heavy chain disulphide bridges, whereas papain splits at the N-terminal side of the inter-heavy chain disulphide bridges. We report here evidence that this model does not apply to all IgG subclasses. In the case of human IgG₂ subclass myeloma proteins, papain splits initially at the C-terminal side of inter-heavy chain disulphide bridges. We also show that the amino-acid sequence of the Fc fragment of human IgG₂ subclass so far determined has approximately 95% homology with that of human IgG₁ and IgG₄ subclasses reported by others^9–15.