B-Cell superantigens: Definition and potential impact on the immune response

Superantigens have been extremely helpful tools in exploring fundamental questions in immunobiology including mechanisms of cell activation, tolerance, and autoimmunity. Until recently, attention has been focused exclusively on T-cell superantigens. However, new data suggest that there are superantigens that directly activate B cells. By definition, these agents (1) stimulate a high frequency of B cells, (2) target B cells that have restricted usage of VH or VL family genes, and (3) bind to immunoglobulins outside the sites that bind conventional antigens. A candidate B-cell superantigen that has received considerable attention in this laboratory is staphylococcal protein A. This agent is best known to the immunologist because of its ability to bind to the Fc fragment of IgG. This binding has been localized to twoα-helical structures on each of four or five homologous regions that comprise the extracellular domain of protein A. However, it is now clear that protein A contains a second site that binds to determinants on the Fab regions of certain immunoglobulins independently of their heavy-chain isotype. In man this so-called alternative site appears to bind only to immunoglobulins that utilize heavy-chain genes of the VH3 subfamily. In the mouse this type of binding is restricted to immunoglobulins using heavy chains belonging to the S107 and J606 VH families. In this review, we examine the growing list of microbial products that dominate B-cell superantigenic properties. Using staphylococcal protein A as a model for a B-cell superantigen, we consider the potential impact of this novel class of antigens on the immune response. We focus on the ability of B-cell superantigens to influence the expression of the B-cell repertoire. In addition, we consider the hypothesis that the interaction of a B-cell superantigen with its reactive serum immunoglobulins activates the classical complement cascade and thus represents a powerful stimulant of tissue inflammation.     

The murine clan V(H) III related 7183, J606 and S107 and DNA4 families commonly encode for binding to a bacterial B cell superantigen

Superantigens, by virtue of their unconventional binding interactions with Ag receptors, can simulate a large subset of mature lymphocytes in the repertoire. Recent studies have documented that in vivo exposure to the model bacterial B cell superantigen, Staphylococcal protein A (SpA), induces large scale effects on murine B-cell clonal selection by mechanism(s) that include deletion of supra-clonal sets. While the structural bases for the immunomodulatory properties of several T-cell superantigens have been well characterized, the requirements for murine Fab-binding of SpA remain incompletely defined. To investigate these structural requirements, a series of direct binding and inhibition studies were performed with a large panel of Moabs of diverse variable region gene usage. These studies confirm previous reports that superantigen binding is completely restricted to the products of clan V(H) III-related families, that include the small S107 and J606 families, and we also demonstrated that usage of the related small DNA4 family commonly correlates with weaker binding activity. Furthermore, our results document that genes from the largest clan V(H) III family, 7183, commonly encode for Fab-mediated binding of SpA, while antibodies from five other VH families, J558, Q52, Sm7, VH11 and VH12, did not display Fab-mediated SpA binding activity. By contributing to the essential foundation for understanding of the structural basis for binding interactions, these findings will aid interpretation of evolving observations regarding the clonal fates induced by in vivo B-cell superantigen exposure.     

Analysis of human monoclonal antibodies elicited by vaccination with a Cryptococcus neoformans glucuronoxylomannan capsular polysaccharide vaccine.

The Cryptococcus neoformans capsular polysaccharide glucuronoxylomannan (GXM) has been conjugated to tetanus toxoid (GXM-TT) as an investigational vaccine. GXM-TT elicits antibodies that are protective in C. neoformans-infected mice. In an effort to characterize the fine specificity and molecular structure of human GXM-TT-elicited antibodies, we generated two GXM monoclonal antibodies (MAbs) from peripheral blood lymphocytes of a volunteer GXM-TT recipient and studied serum GXM antibody idiotype expression in 10 additional vaccinees. The MAbs, 2E9 and 3B6, are the immunoglobulin M(lambda) isotype and bind capsular polysaccharides of C. neoformans serotypes other than the serotype A that was used for immunization. Neither antibody competes with murine GXM MAbs for antigen binding, suggesting that the human MAbs recognize a different epitope. The B-cell superantigen staphylococcal protein A binds both MAbs, and human immunodeficiency virus gp120 binds 2E9. MAb nucleic acid sequence analysis revealed that both antibodies use an identical V lambda 1a-J lambda genetic element with different, somatically mutated, members of the VH3 gene family and different DH and JH gene elements. The gene elements used by both MAbs occur in fetal B-lymphocyte repertoires, autoantibodies, and other polysaccharide antibodies. Post-GXM-TT vaccination GXM antibodies from 10 additional vaccinees expressed a shared idiotype defined by rabbit antiserum raised against MAb 2E9. Our data suggest that the human GXM antibody response is restricted and raise questions regarding the importance of specific variable-region elements and superantigens in the generation of human antibody responses to encapsulated pathogens.     

Monoclonal anti-VH141 antibodies that specifically recognize the heavy chain variable region of, and are closely related to, MOPC141 myeloma protein whose VH gene belongs to VHQ52 family.

To raise monoclonal antibodies that specifically recognize the heavy chain variable region of MOPC141 myeloma protein (VH141), which belongs to VHQ52 family, rats were immunized with Fd'-conjugated keyhole limpet haemocyanin (KLH) (Fd': Fd' fragments of MOPC141), and the spleen cells were fused with mouse myeloma cells. The resulting 900 hybridomas were screened for antibody activity against Fd'1 fragments having no constant H-chain sequences, which were prepared by cleavage of the Fd' fragments with cyanogen bromide, and two monoclonal antibodies, designated 3-2-7h and 3-5-6f, were obtained. Radioimmunoassay inhibition test showed that the two monoclonal antibodies specifically recognized the VH141, but each was directed to a different determinant on the VH141. When the functional VH gene of Abelson virus-transformed mu-producing pre-B cells, which could be strongly stained with 3-5-6f monoclonal antibody, was cloned and sequenced, the VH gene was closely relate to that of MOPC141 (88% and 94% homology at amino acid and DNA level, respectively). Taken together, the results indicated that 3-2-7h had high specificity only for the VH141, whereas 3-5-6f specifically reacted not only with the VH141 but also with the VH region closely related to that of MOPC141, and that both the monoclonal anti-VH141 antibodies were specific for a limited range of VH regions within the VHQ52 family rather than being VHQ52 family specific. These monoclonal anti-VH141 antibodies should be very useful to determine at a single cell level by immunofluorescence the usage of the VH gene(s) identical or closely related to that of MOPC141 during early B-cell development.     

Characterisation of a mouse monoclonal anti-idiotype reactive with a V region sequence commonly used by human immunoglobulins.

BACKGROUND: A mouse monoclonal antibody (2C7/IgG2b kappa) has been described recently, which is directed against the major house dust mite allergen Der p 1, and whose epitope specificity is representative of a major component of the human IgE anti-Der p 1 response. AIMS: To characterise an anti-idiotypic antibody (2G10/IgG1 kappa) raised against monoclonal antibody 2C7 as surrogate human IgE anti-Der p 1. METHODS: The specificity of the anti-idiotype antibody 2G10 was determined by competitive inhibition experiments using human and mouse immunoglobulins of known VH gene families. The epitope recognised by monoclonal antibody 2G10 was located on the molecular model of the Fv (fragment variable) region of monoclonal antibody 2C7. RESULTS: The data suggest that monoclonal antibody 2G10 is directed against a crossreactive idiotype on human IgE that is shared by polyclonal IgG. Competitive inhibition studies against human immunoglobulins, representative of VH2, VH3, and VH4 gene families, showed that monoclonal antibody 2G10 is mostly likely to be directed against sequences encoded by either VH3 or VH4 genes. The fact that monoclonal antibody 2G10 binds to the humanized (complementarity determining region (CDR) grafted) CAMPATH-1H antibody, but not to the original rat CAMPATH-1 YTH34.5.6 antibody, indicates that it is directed against a framework region rather than the CDRs. Analysis of amino acids in the VH region for charge, hydrophobicity, and accessibility suggests that reactivity with monoclonal antibody 2G10 is defined by a hexapeptide spanning residues 74-79 within framework region 3. CONCLUSION: The anti-idiotype monoclonal antibody 2G10 could potentially be used as a probe for determining the contribution of the VH3 and VH4 gene segments to antigenic specificity.     

Staphylococcus aureus Cowan I-induced human immunoglobulin responses: Preferential IgM rheumatoid factor production and VH3 mRNA expression by protein A-binding B cells

Protein A (PA), a cell wall component of SAC, activates human B cells by cross-linking the Fabs of membrane immunoglobulins. Recent data indicate that PA binds only to Fabs that use VH3 heavy chains, and thus it has been designated as a B-cell superantigen. We previously reported thatStaphylococcus aureus Cowan I (SAC) -induced IgM rheumatoid factor (RF) by human PBMC was mediated by PA. Therefore, we sought to determine if SAC-induced IgMRF production was confined to PA-binding B cells and if these B cells were enriched for the expression of VH3 heavy chains. We observed that the elicitation of IgMRF in response to SAC was limited to a subset of B cells that bind PA and that this subset was enriched for VH3 mRNA expression. Taken together, these results suggest that IgMRFs produced in response to SAC are enriched for usage of VH3 heavy chains. Thus, this SAC-induced autoantibody response appears to represent a new B-cell superantigenic property of PA.     

Immunoglobulin heavy-chain receptor editing is observed in the NOD/SCID model of human B-cell development

Receptor editing and receptor revision are the two mechanisms of antibody diversity that result in either complete V-gene replacement or the formation of hybrid V genes. We do not yet understand how this process unfolds, because they are rare and difficult to study in vivo. In this study, we describe a family of VH4-34:VH4-61 hybrids isolated from a human B-cell chimeric non-obese diabetic/severe combined immunodeficient mouse. The observation of hybrid immunoglobulin sequences in human B cells that developed in this model system makes it useful for the study of this mechanism of diversification and tolerance.     

Anti-idiotypic antibody D12 and superantigen SPA both interact with human VH3-encoded antibodies on the external face of the heavy chain involving FR1, CDR2 and FR3

The mouse monoclonal antibody (mAb) D12 specifically binds in the variable region (idiotype) of human V_H3 encoded antibodies. We used mutational analysis to determine the subregions of a V_H3 encoded antibody which effect the interaction with mAb D12. Recombinant antibodies composed of mutant heavy chains were produced using the baculovirus expression system. The results of this topographical study indicate that the combined conformations of FR1, CDR2 and FR3 are critical for mAb D12 binding. MAb D12 binding was not effected either by the heavy chain CDR3 sequence nor by the light chain. We previously demonstrated that structures within the same three subregions are required for the B cell superantigen Staphylococcal protein A (SPA) binding to V_H3 encoded antibodies. Thus, some anti-idiotypic antibodies can interact with antibodies in a similar fashion to superantigens.     

Monoclonal BALB/c anti-progesterone antibodies use family IX variable region heavy chain genes

Variable region nucleotide sequences and respective translated amino acid sequences for three heavy chains (DB3, 11/32 and 10/8) and two light chains (DB3 and 11/32) of monoclonal mouse IgG1 anti-progesterone antibodies have been determined by primer extension mRNA sequencing. The three VH regions exhibit the same rarely observed VH IX gene family and have greater than 88% homology between them. Two associated light chain sequences are 95% homologous and belong to the V kappa I group. The N-terminal twenty two amino acids of the kappa light chain of the third antibody 10/8 have been determined by automated protein sequencing and are identical to those of 11/32. Thus, these three monoclonal anti-progesterones derived from separate fusions all use VHIX-V kappa 1 gene combinations.     

Clonal analysis of a human antimouse antibody (HAMA) response

Circulating human antimouse antibodies (HAMAs) directed to mouse immunoglobulin G (IgG) are clinically significant, compromising mouse antibody therapy and imaging, and interfering in immunological assays. To investigate the HAMA response, 20 stable cell lines secreting human monoclonal antibodies reactive with mouse IgG were established from a donor with a history of exposure to mice. Their subclass and domain specificities were established by solid-phase binding, indirect haemagglutination assays and immunoblotting, using Igs of known subclass and Ig fragments. The heavy-chain variable region gene usage was determined for 12 HAMAs. Eight HAMAs were IgM, 11 HAMAs were IgG4 and one HAMA was IgG1, indicating an IgG4-dominated response. All of the IgG HAMAs reacted with epitopes present on the Fc portion; one was subclass-specific, nine were subclass-restricted and two were pan-IgG-reactive. Measurement of their affinities gave dissociation constants typically in the nanomolar range. Seven and five HAMAs were derived from variable heavy-chain 3 (VH3) and VH1 gene segments, respectively. The IgG HAMAs used different VH segments to the IgM HAMAs. JH regions were coded by JH4 in eight HAMAs. DH segment usage appeared to be restricted in the IgM HAMAs. Two IgG HAMAs were clonally related. These monoclonal HAMAs are potentially useful as reagents for detecting mouse IgG and as reference reagents for the investigation of the HAMA response in patients undergoing mouse monoclonal antibody therapy and for the investigation of the influence of HAMAs on immunodiagnostic tests.     

Molecular analysis of carbohydrate antigen-induced monoclonal IgM anti-IgG antibodies (rheumatoid factors).

Light and heavy chain variable regions of 11 monoclonal rheumatoid factors (MRF) produced after carbohydrate antigen immunization, and one MRF produced after protein immunization have been sequenced. Most carbohydrate antigen induced MRF utilized light chains that were homologous to light chains of MRF obtained from protein immune or LPS stimulated mice, and MRF derived from the autoimmune MRL/lpr mouse strain. VH gene usage was diverse for carbohydrate antigen induced MRF that bound all four isotypes of IgG, or that bound only the IgG3 isotype. In contrast VH gene use among our panel of MRF that bound the IgG1 isotype appeared restricted. Four of the five IgG1 binders used VH genes that were highly homologous to the VH nucleotide sequence of a gene encoding an NP binding monoclonal antibody. Our study confirms the use of a particular group of light chain genes among murine MRF, confirms that there is diversity in the heavy chain genes utilized among MRF, and suggests that a gene(s) homologous to the VH NP 23 J558 gene may be preferentially associated with murine MRF specificity for the IgG1 isotype.     

Somatically mutated IgG anti-DNA antibody clonally related to germ-line encoded IgM anti-DNA antibody.

Among 15 anti-DNA antibody-producing hybridomas derived from a single NZB X NZW F1 mouse, an IgM and an IgG were shown to use the same VH gene of the Q52 family. Using a combination of two primers both consisting of a mixture of oligonucleotides (one complementary to the 5' end of VH segment and one to the 3' end of VH segment of Q52 family) we determined the sequences of several members of germ-line VH genes in the Q52 family derived from NZB and NZW strains. Comparison of the sequences with those of cloned VH cDNA obtained from the hybridomas revealed that the VH sequence of the IgM anti-DNA antibody was identical to that of a cloned NZW germ-line VH gene, except for the priming sites. In contrast, the VH sequence of the IgG counterpart contained somatically mutated nucleotides. Because the IgG anti-DNA antibody showed a higher DNA binding activity than did the IgM antibody, we conclude that these changes in nucleotide sequences were induced and selected through an antigen-driven mechanism as is the case in a normal immune response. It is tempting to speculate that the germ-line encoded, low-affinity IgM autoantibody undergoes somatic mutations and isotype switching, resulting in generation of pathogenic, high-affinity autoantibodies in autoimmune diseases.     

High efficiency creation of human monoclonal antibody-producing hybridomas

The native human antibody repertoire holds unexplored potential for the development of novel monoclonal antibody therapeutics. Current techniques that fuse immortal cells and primary B-lymphocytes are sub-optimal for the routine production of hybridomas that secrete human monoclonal antibodies. We have found that a murine cell line that ectopically expresses murine interleukin-6 (mIL-6) and human telomerase (hTERT) efficiently forms stable human antibody-secreting heterohybridomas through cell fusion with primary human B-lymphocytes. The hybrid cells maintain secretion of human antibodies derived from the primary B-lymphocytes through multiple rounds of cloning. Using splenic B-lymphocytes from a patient immunized with a Streptococcus pneumoniae capsular polysaccharide vaccine, we have succeeded in creating hybridomas that secrete human monoclonal antibodies specific for S. pneumoniae antigens. Using peripheral blood lymphocytes, we have similarly cloned a human antibody that binds a viral antigen. These experiments establish that SP2/0-derived cell lines ectopically expressing mIL-6 and hTERT will enable the rapid cloning of native human monoclonal antibodies.     

PATHOGENICITY AND SEQUENCE ANALYSIS OF A MOUSE MONOCLONAL ANTIBODY AGAINST HUMAN ACETYLCHOLINE RECEPTOR IN MYASTHENIA GRAVIS

通过对乙酰胆碱受体（ＡＣｈＲ）自身抗体分子结构以及与致病性关系的研究探讨重症肌无力（ＭＧ）及其动物模型——实验性自身免疫性重症肌无力（ＥＡＭＧ）的发病机理。ＡＣｈＲ抗体被动转移至大鼠后诱导出明显的ＥＡＭＧ。全身肌肉ＡＣｈＲ损失率和体重减轻率达４７．２±１５．３％和１３．４±２．２％。这株ＡＣｈＲ抗体的重链可变区基因由小鼠Ｑ５２胚系基因编码，其同源性为９４．８％，将这株抗体的重链和轻链可变区、尤其是互补决定区（ＣＤＲ）的核苷酸和氨基酸序列与其他致病性ＡＣｈＲ抗体比较发现，能诱导ＭＧ和ＥＡＭＧ的致病性ＡＣｈＲ抗体的结构并不是完全一致的。     

Perspectives on Antigenicity and Idiotypy

HIV-1 infection affects many of the cellular components vital for the maintenance of immune homeostasis. Similar to the T cell superantigen effect on T cell expansion and depletion in AIDS, HIV components with B cell superantigenic properties could be responsible for the observed B cell activation and skewing of VH family usage. Current data on possible B cell superantigen properties of HIV proteins (gp120) are mostly based on studies describing the clonality and VH family usage of immunoglobulins in HIV infection. Various laboratories reported independently an unusual skewing of the VH-repertoire of antibodies that appears not to be random. According to these observations, an enrichment of VH 1 and VH4 family - paralleled a depletion of VH3 family - utilizing anti-HIV-1 gp120 and p24 antibodies in HIV-1 infected individuals and a loss of total VH3⁺ Ig in patients with late stages of AIDS. Polyclonal and monoclonal (VH1, VH4, and VH5) anti-p24 and gp120 antibodies share a crossreactive idiotype (IF7). IF7 like antibodies were found in the serum of HIV-1 infected individuals, persisting in the course of infection, perhaps contributing to the depletion of VH3⁺ Ig. Furthermore, a restriction of clonal heterogeneity of anti-p24 and anti-gp120 antibodies was detected by isoelectric focusing and indicated by skewed kappa/lambda light chain isotype ratios, indicating clonal dominance of certain sets of anti-HIV-1 antibodies during infection. Taken these findings together, a strong case for the involvement of a B cell superantigen can be made, although the mechanism of B cell depletion is not fully understood.     

Immunoglobulin Gene Mutations and Frequent Use of VH1-69 and VH4-34 Segments in Hepatitis C Virus-Positive and Hepatitis C Virus-Negative Nodal Marginal Zone B-Cell Lymphoma

Nodal marginal zone B-cell lymphoma (NMZL) is actually considered as a distinct entity that must be distinguished from extra-nodal and splenic marginal zone lymphomas. To define the cell origin and the role of antigen stimulation we determined the nucleotide sequence of the tumor-related immunoglobulin heavy chain variable genes in 10 cases of NMZL. The results were also evaluated on the basis of the presence of chronic hepatitis C virus (HCV) infection. All 10 cases harbored VH somatic mutations with a sequence homology compared to the closest germline gene, ranging from 83.33 to 98.28%. Interestingly, different VH segments were preferentially used in HCV-positive and HCV-negative patients: three of five HCV-negative NMZLs used a VH4-34 segment joined with different D and JH segments whereas three of five HCV-positive NMZLs used a VH1-69 gene joined with a D3-22 and a JH4 segment, with very strong similarities in the CDR3s among the three different cases. These data indicate: 1) NMZL is derived from B cells that have experienced the germinal center reaction; 2) the preferential usage of a VH1-69 segment in the majority of the HCV-positive NMZL cases with similar CDR3s suggests the presence of a common antigen, probably a HCV antigen epitope, involved in the B-cell selection; and 3) the use of a VH4-34 segment suggests a role of yet unknown B-cell superantigen(s) in the selection of tumor B-cell precursors in HCV-negative NMZL.     

The molecular origin of anti-DNA antibodies

The in vitro observation that a single point mutation in the protective anti-phosphorylcholine anti-bacterial antibody, S107, converts it into an autoantibody that reacts with dsDNA has focused our attention on the role of somatic mutation in generating autoantibodies. It has also led us to examine the significance of an individual's prior response to environmental antigens on the subsequent production of autoantibodies. The fact that genes of the S107 heavy chain variable region family could encode autoantibodies made it possible to clone and sequence the relevant germline genes of this small family from autoimmune (NZB x NZW)F1 mice and to compare these to the comparable genes in non-autoimmune mice. The germline genes from the normal and autoimmune mice are quite homologous and the small number of polymorphisms are not likely to predispose the autoimmune mice to the production of autoantibodies. (NZB x NZW)F1 mice respond to immunization with phosphorylcholine with a response that is largely encoded by the VH1 gene of the S107 family. However, when these same mice begin to make autoantibodies, their anti-DNA antibodies which are encoded by this family are in fact derived from the VH11 gene. The VH11 encoded anti-DNA antibodies which have been sequenced are all of the IgG2a subclass, react with dsDNA, and have undergone significant somatic diversification from the germline gene. Analysis of the ratio and location of the replacement and silent mutations suggests that the regulation of the autoantibody response differs from that of the normal response to foreign antigens. Our studies suggest that the utilization of a particular VH germline gene in the immune response to foreign antigens early in life does not lead to the preferential utilization of that same gene in the subsequent production of autoantibodies.     

Comparative study of VH gene family usage by newborn xid and non-xid mice, newborn NZB and adult NZB mice, and by splenic and peritoneal cavity B cell compartments

RNA from 170 different Ig-secreting hybridomas was hybridized with VH gene probes 7183, QUPC52, S107, J558, J606, 36-60, V31, X-24 and V-GAM 3.8, with the aim of comparing xid to non-xid mice, neonatal NZB to adult NZB, and the peritoneal cavity with the splenic compartments for VH gene family expression. Our results indicate that (a) defective F1 male xid mice express with high frequency 36-60 and J606 5' proximal VH families, when compared to non-xid F1 females of the same matings. (b) Neonatal NZB mice express 3' proximal 7183 and QUPC52 families with high frequency, when compared to adult mice and to the expected values derived from VH complexity. (c) Natural antibodies against cytoskeleton proteins, DNA, rabbit IgG and 2,4,6-trinitrophenyl did not appear to prefer a particular VH family. The only difference found is related to age; neonatal clones preferentially employ 7183 and QUPC52 while J558 predominates among adult clones. (d) All the 15 antibodies directed against bromelin-treated mouse red blood cells failed to hybridize with any of the nine VH probes employed. These results confirm previous findings indicating the highly homogeneous pattern of these latter antibodies, and suggest that they are encoded by a new VH family (VH11).