Molecular approaches to the study of human B-cell and (auto)antibody repertoire generation and selection

Differences in the composition of the human antibody repertoire by B cell subsets in the blood

The vast initial diversity of the antibody repertoire is generated centrally by means of a complex series of V(D)J gene rearrangement events, variation in the site of gene segment joining, and TdT catalyzed N-region addition. Although the diversity is great, close inspection has revealed distinct and unique characteristics in the antibody repertoires expressed by different B cell developmental subsets. In order to illustrate our approach to repertoire analysis, we present an in-depth comparison of V(D)J gene usage, hydrophobicity, length, D_H reading frame, and amino acid usage between heavy chain repertoires expressed by immature, transitional, mature, memory IgD⁺, memory IgD⁻, and plasmacytes isolated from the blood of a single individual. Our results support the view that in both human and mouse, the H chain repertoires expressed by individual, developmental B cell subsets appear to differ in sequence content. Sequencing of unsorted B cells from the blood is thus likely to yield an incomplete or compressed view of what is actually happening in the immune response of the individual. Our findings support the view that studies designed to correlate repertoire expression with diseases of immune function will likely require deep sequencing of B cells sorted by subset.

DH and JH usage in murine fetal liver mirrors that of human fetal liver

In mouse and human, the regulated development of antibody repertoire diversity during ontogeny proceeds in parallel with the development of the ability to generate antibodies to an array of specific antigens. Compared to adult, the human fetal antibody repertoire limits N addition and uses specifically positioned VDJ gene segments more frequently, including V6-1 the most D(H)-proximal V(H,) DQ52, the most J(H)-proximal D(H), and J(H)2, which is D(H)-proximal. The murine fetal antibody repertoire also limits the incorporation of N nucleotides and uses its most D(H) proximal V(H), V(H)81X, more frequently. To test whether D(H) and J(H) also follow the pattern observed in human, we used the scheme of Hardy to sort B lineage cells from BALB/c fetal and neonatal liver, RT-PCR cloned and sequenced V(H)7183-containing VDJCμ transcripts, and then assessed V(H)7183-D(H)-J(H) and complementary determining region 3 of the immunoglobulin heavy chain (CDR-H3) content in comparison to the previously studied adult BALB/c mouse repertoire. Due to the deficiency in N nucleotide addition, perinatal CDR-H3s manifested a distinct pattern of amino acid usage and predicted loop structures. As in the case of adult bone marrow, we observed a focusing of CDR-H3 length and CDR-H3 loop hydrophobicity, especially in the transition from the early to late pre-B cell stage, a developmental checkpoint associated with expression of the pre-B cell receptor. However, fetal liver usage of J(H)-proximal D(H)Q52 and D(H)-proximal J(H)2 was markedly greater than that of adult bone marrow. Thus, the early pattern of D(H) and J(H) usage in mouse feta liver mirrors that of human.

Immune responses of teleost fish

In fish all the pre-requisites to mount a specific immune response are present, but the main differences from the mammalian system are that the secondary response is relatively minor and IgG is not present. In teleosts mainly IgM is present, and IgD has been recently described but its function is, as yet, unknown. However, different forms of fish IgM and its observed flexibility of structure may compensate for a lack of Ig class diversity. The innate immune response of teleosts is highly developed. Multiple forms of key constitutive and inducible components, such as lysozyme, C3, alpha2-macroglobulin and C-reactive protein, are present, and may enhance immune recognition. Low ambient temperature appears to have an impact on all aspects of the immune response, particularly the T-dependent specific immune response due to the non-adaptive lipid composition of T-cell membranes. Temperature effects on the nonspecific immune system are less well characterised, but there is evidence that low temperatures are also suppressive. Knowledge of immune system function becomes essential for disease prevention strategies such as the development of vaccines, selection for increased disease resistance and identification of genes suitable for trangenesis.

Immunoglobulin VH-gene usage of autoantibodies in mercuric chloride-induced membranous glomerulopathy in the rat

Brown-Norway (BN) and Dorus Zadel Black (DZB) rats develop a T-cell-dependent membranous glomerulopathy (MGP) with high proteinuria and antiglomerular basement membrane (GBM) autoreactive antibodies (Abs), upon exposure to mercuric chloride (HgCl2). Laminin is an important autoantigenic target of the anti-GBM Abs, absorbing approximately 30% of the anti-GBM reactivity. Although many anti-GBM Abs have undergone isotype switching, it is currently unclear whether affinity maturation occurs during the HgCl2-induced autoimmune response. To address this question we analysed the rearranged immunoglobulin heavy chain variable-region genes (VHDJH regions) of 15 mAbs that were previously obtained from HgCl2-treated rats. Seven of these mAbs exhibit reactivity towards laminin. Our study showed that the VH-gene usage of antilaminin mAbs is largely restricted to the PC7183 VH-gene family (six out of seven). In addition, we demonstrated that at least three out of six laminin reactive and five out of six non-laminin-binding mAbs are encoded by germline VH genes (a total of eight out of 12 mAbs). Of the eight mAbs that are encoded by germline VH genes, seven are of a non-immunoglobulin M (IgM) isotype, indicating that isotype switching has occurred in these mAbs in the absence of somatic mutations. The mutations observed in the VH genes of the four remaining mAbs do not provide strong evidence for antigenic selection. The data support the notion that B cells in this model of MGP are not subjected to affinity maturation and probably result from polyclonal B-cell activation.

Changing the antigen binding specificity by single point mutations of an anti-p24 (HIV-1) antibody

The murine mAb CB4-1 raised against p24 (HIV-1) recognizes a linear epitope of the HIV-1 capsid protein. Additionally, CB4-1 exhibits cross-reactive binding to epitope-homologous peptides and polyspecific reactions to epitope nonhomologous peptides. Crystal structures demonstrate that the epitope peptide (e-pep) and the nonhomologous peptides adopt different conformations within the binding region of CB4-1. Site-directed mutagenesis of the fragment variable (Fv) region was performed using a single-chain (sc)Fv construct of CB4-1 to analyze binding contributions of single amino acid side chains toward the e-pep and toward one epitope nonhomologous peptide. The mutations of Ab amino acid side chains, which are in direct contact with the Ag, show opposite influences on the binding of the two peptides. Whereas the affinity of the e-pep to the CB4-1 scFv mutant heavy chain variable region Tyr(32)Ala is decreased 250-fold, the binding of the nonhomologous peptide remains unchanged. In contrast, the mutation light chain variable region Phe(94)Ala reduces the affinity of the nonhomologous peptide 10-fold more than it does for the e-pep. Thus, substantial changes in the specificity can be observed by single amino acid exchanges. Further characterization of the scFv mutants by substitutional analysis of the peptides demonstrates that the effect of a mutation is not restricted to contact residues. This method also reveals an inverse compensatory amino acid exchange for the nonhomologous peptide which increases the affinity to the scFv mutant light chain variable region Phe(94)Ala up to the level of the e-pep affinity to the wild-type scFv.

Analysis of human self-reactive antibody repertoires by quantitative immunoblotting

We review the use of a quantitative immunoblotting technique to characterize human self-reactive antibody repertoires in health and disease. The interactions of plasma IgM and IgG with tissue extracts as sources of self-antigens were analyzed by quantitative immunoblotting. Data were compared by means of multiparametric statistical analysis. The data summarized here demonstrate that natural self-reactive antibody repertoires of healthy individuals are restricted to a limited subset of immunodominant autoantigens that is selected early in development, and remains conserved between individuals through ageing. The selection of human natural self-reactive IgG antibody repertoires requires normal T-/B-cell interactions. The immunoblotting assay has the potential to distinguish between autoimmune diseases with organ-related oligoclonal expansion of self-reactive clones and those characterized by broad alterations of immunoregulation. However, organ-specific autoimmune diseases may be characterized by altered patterns of antibody repertoires unrelated to the target organ. The assay also revealed an unexpected defect in the regulatory function of self-reactive IgM on the expression of self-reactive IgG repertoires in several systemic and organ-specific autoimmune diseases. The results are discussed in the light of our current understanding of the processes of selection of self-reactive B-cells and the pathophysiology of autoimmunity.

Molecular Mechanisms and Selection Influence the Generation of the Human VλJλ Repertoire

To define the λ light chain repertoire in humans, a single-cell PCR technique using genomic DNA obtained from individual peripheral B cells was employed. Of the 30 known functional Vλ genes, 23 were detected in either the nonproductive or productive repertoires. Specific Vλ genes, including 2A2, 2B2, 1G, and 4B, were overexpressed in the nonproductive repertoire, whereas some Vλ genes, such as 3R, 2A2, 2B2, 1C, 1G, and 1B, were overexpressed in the productive repertoire. Comparison of the nonproductive and productive repertoires indicated that no Vλ genes were positively selected, whereas a number of Vλ genes, including 4C, 1G, 5B, and 4B, were negatively regulated. All four of the functional Jλ segments were found in both repertoires, with Jλ7 observed most often. Evidence of terminal deoxynucleotidyltransferase activity was noted in nearly 80% of nonproductive VλJλ rearrangements, and exonuclease activity was apparent in the majority. Despite this, the mean CDR3 length was 30 base pairs in both productive and nonproductive repertoires, suggesting that it was tightly regulated at the molecular level. These results have provided new insights into the dimensions of the human Vλ repertoire and the influences that shape it.

Similar characteristics of the CDR3 of V(H)1-69/DP-10 rearrangements in normal human peripheral blood and chronic lymphocytic leukaemia B cells

The variable heavy chain (V(H)) gene segment V(H)1-69/DP-10 has been shown to be over-represented in B-cell chronic lymphocytic leukaemia (CLL). Because of certain similar characteristics of their complementarity determining region 3 (CDR3), including preferential utilization of J(H)6 elements and an extended length, it has been suggested that antigenic stimulation might be involved in leukaemogenesis. Utilizing single-cell PCR to amplify and sequence genomic DNA from individual normal human peripheral blood B cells, we have obtained 7/421 productively and 1/69 nonproductively rearranged V(H) genes that used V(H)1-69/DP-10. All productive rearrangements were unmutated, used J(H)6 and had an average CDR3 length similar to that previously found in V(H)1-69/DP-10-expressing CLL cells. These results suggest that CLL may arise from B cells commonly found in the peripheral B-cell repertoire and do not represent expansion of a unique subset of specific antigen-reactive B cells.

Analysis of the human VH gene repertoire. Differential effects of selection and somatic hypermutation on human peripheral CD5(+)/IgM+ and CD5(-)/IgM+ B cells

To analyze the immunoglobulin repertoire of human IgM+ B cells and the CD5(+) and CD5(-) subsets, individual CD19(+)/ IgM+/CD5(+) or CD5(-) B cells were sorted and non-productive as well as productive VH gene rearrangements were amplified from genomic DNA and sequenced. In both subsets, the VH3 family was overrepresented largely as a result of preferential usage of a small number of specific individual family members. In the CD5(+) B cell subset, all other VH families were found at a frequency expected from random usage, whereas in the CD5(-) population, VH4 appeared to be overrepresented in the nonproductive repertoire, and also negatively selected since it was found significantly less often in the productive compared to the nonproductive repertoire; the VH1 family was significantly diminished in the productive rearrangements of CD5(-) B cells. 3-23/DP-47 was the most frequently used VH gene segment and was found significantly more often than expected from random usage in productive rearrangements of both CD5(+) and CD5(-) B cells. Evidence for selection based on the D segment and the JH gene usage was noted in CD5(+) B cells. No differences were found between the B cell subsets in CDR3 length, the number of N-nucleotides or evidence of exonuclease activity. Somatically hypermutated VHDJH rearrangements were significantly more frequent and extensive in CD5(-) compared to CD5(+) IgM+ B cells, indicating that IgM+ memory B cells were more frequent in the CD5(-) B cell population. Of note, the frequency of specific VH genes in the mutated population differed from that in the nonmutated population, suggesting that antigen stimulation imposed additional biases on the repertoire of IgM+ B cells. These results indicate that the expressed repertoire of IgM+ B cell subsets is shaped by recombinational bias, as well as selection before and after antigen exposure. Moreover, the influences on the repertoires of CD5(+) and CD5(-) B cells are significantly different, suggesting that human peripheral blood CD5(+) and CD5(-) B cells represent different B cell lineages, with similarities to murine B-1a and B-2 subsets, respectively.

Monoclonal immunoglobulin A derived from peritoneal B cells is encoded by both germ line and somatically mutated VH genes and is reactive with commensal bacteria

We transferred peritoneal cells from BALB/c mice into C.B17 scid/scid mice. Six to eight months after injection, only cells with the B1 phenotype were retained in the spleens and peritoneal cavities of these mice. The lamina propria of the intestine contained many peritoneal, donor-derived, immunoglobulin A (IgA)-producing cells. The mesenteric lymph nodes of these mice were found to be a major site of proliferation and generation of IgA plasmablasts. We established eight IgA-producing hybridomas from the mesenteric lymph nodes of such mice, and all the hybridomas reacted with different but partially overlapping fecal bacterial populations. Cloning and sequencing of the VH genes of these hybridomas showed that two hybridomas utilized germ line-encoded VH genes while the VH genes of the six hybridomas showed somatic mutations, some of which are indicative of an antigen-driven selection process.

Immunoglobulin variable gene analysis of human autoantibodies reveals antigen-driven immune response to glutamate decarboxylase in type 1 diabetes mellitus

Insulin-dependent (type 1) diabetes mellitus is an organ-specific autoimmune disease frequently associated with an islet-specific humoral autoimmune response. The role of islet cell autoantibodies in the disease process is unclear; in particular, it is not known whether they are a non-specific side effect of islet cell destruction or play a role in the autoimmune network leading to type 1 diabetes. Here we report the immunoglobulin gene usage and somatic mutation rates of a panel of seven human monoclonal islet cell autoantibodies (MICA 1-7) directed towards the major islet cell autoantigen glutamate decarboxylase (GAD). These autoantibodies were produced from cells from two patients with newly diagnosed type 1 diabetes. VH1, VH4 and V lambda 2 gene segments were frequently used in the MICA, but no correlation between V gene usage and epitope recognition was found. The nonrandom ratio of replacement versus silent mutations in the variable gene region, an accumulation of replacement mutations in the complementarity determining regions, which confer antigen binding, and the high relative avidity for GAD observed for MICA 1, 3, 4, and 6, suggested that the immune response to GAD is driven by the antigen. In contrast, MICA 2, 5, and 7, revealing a lower affinity for antigen, have accumulated a large number of silent mutations. These latter antibodies may, therefore, be characteristic for later stages of the chronic autoimmune disease. Our results argue in favor of an antigen-driven autoantibody response to islets in human type 1 diabetes. They suggest that GAD is an important target of autoimmunity associated with type 1 diabetes.

The immunoglobulin VH gene, VH4-21, specifically encodes autoanti-red cell antibodies against the I or i antigens

Most autoanti-red cell antibodies found in patients with cold agglutinin disease are specific for the I or i carbohydrate antigenic determinants. However, antibodies specific for other antigens such as Pr or Sa can also be found, and these are identified by their pattern of reactivity with enzyme-treated red cells. Recently, it has been shown that the vast majority of anti-Ii antibodies react with a monoclonal anti-idiotypic antibody (9G4); this reactivity arises from restriction of the immunoglobulin heavy chains used to encode the antibodies to a single VH4-21 gene, VH4-21. The 9G4 antibody appears specific for this gene product, and we have used it to analyse VH4-21 gene involvement in encoding a spectrum of red cell antibodies of various specificities. The results support the strong association between usage of this gene and anti-Ii specificity and indicate that it is not generally used by other specificities. In particular, it is striking that the unsubstituted type 2 oligosaccharide antigens (I and i) induce a highly restricted autoantibody response very different from that induced by the sialylated type 2 antigens (Sia-b, -1 and 1b). The 9G4 antibody therefore provides a simple tool for discrimination between these autoanti-red cell antibodies, which should be of use in red cell serology.

Differential usage of an autoantibody-associated VH gene, VH4-21, by human B-cell tumors

Selection of immunoglobulin variable region genes for recombination in B cells takes place from among those VH and VL gene segments available in the unrearranged germ line repertoire. In the case of neoplastic B cells, there is apparent deviation in the use of V-genes from that expected on a random basis, both for VH and for VL. Also, the preferred V-genes, and their patterns of mutation, differ among the various categories of B-cell tumor possibly reflecting the distinct origins and clonal histories on the individual tumor cells. This review focuses on a single VH gene, VH4-21, which is a member of the VH4 family, and which appears selectively to encode immunoglobulins with autoantibody activity, particularly anti-red cell antibodies. The pattern of usage of this VH gene by B-cell tumors demonstrates clear asymmetry among different tumor types. Also, the mutations detected in this relatively non-polymorphic gene indicate that antigen, possibly autoantigen, may influence the behavior of the tumor cell.

A reduced level of multiple mutation in a shuttle vector passaged in Sjögren's syndrome cells

Sjögren's syndrome is a systemic disorder with unknown etiology, displaying many signs of autoimmunity. Although the basic mechanism of this disease is unknown, a defect in somatic mutagenesis of antibody genes has been suggested. Using a shuttle vector plasmid, we here show that the number of vectors with multiple base changes in a marker gene was reduced in B cell lines from two patients with Sjögren's syndrome (8% in both), as compared with values reported for cell lines from normal human donors (16-27%). This finding suggests that a reduction of the rate of somatic mutagenesis may influence the development of symptoms in Sjögren patients.

Somatically mutated member of the human V lambda VIII gene family encodes anti-myelin-associated glycoprotein (MAG) activity

A highly conserved small family of human V lambda genes was identified by DNA homology to a V lambda gene isolated from a patient with demyelinating peripheral neuropathy, and which encodes an autoantibody with anti-MAG activity. Comparison of the genes indicates that the patient V lambda gene was derived from one of the germline genes. Together with published analyses of other anti-MAG IgM antibodies, which also appear to be mutated in comparison to known germline V genes, these results suggest that development of these pathogenic antibodies may reflect an antigen-driven, T cell-dependent process.

Light and heavy chains specifying a human IgM kappa autoantibody to a T-cell receptor V beta-antigen

Humans frequently produce serum IgM autoantibodies reactive with T-cell receptor beta chains at a determinant defined by peptides corresponding to the first complementarity determining region. It is likely that this determinant serves as a public idiotype involved in immunoregulation. Following screening of culture supernatants from over 60 Epstein-Barr virus-carrying B-cell lines of normal and neoplastic origin, we identified a line, IARC307, that secretes an IgM kappa protein showing marked specificity for the V beta 8.1 CDR1 sequence CKPISGHNSLFQWYRQT. We cloned and sequenced the complete variable regions of the V kappa and VH chains used by the autoantibody. The light chain has a V kappa III sequence related to the 'a' subgroup and uses J kappa 2. The heavy chain has a VHIII sequence essentially identical to the germ-line sequence DP54 and uses the JH6C minigene. The CDR3 is unique, differing from those of other autoantibodies. The antibody is rigorously specific in its specificity for the V beta 8 peptide and does not show polyspecificity for protein or DNA antigens.

Characterization of a human monoclonal autoantibody directed to cardiolipin/beta 2 glycoprotein I produced by chronic lymphocytic leukaemia B cells

We determined the specificity and sequence of immunoglobulin molecules synthesized by monoclonal B cells from a patient with chronic lymphocytic leukaemia (CLL) who presented with a number of clinical and biological autoimmune symptoms. Heterohybrids obtained by fusion of CLL cells with the mouse X63-Ag 8.653 myeloma produced IgM lambda MoAbs directed to the cardiolipin/beta 2 glycoprotein I (beta 2GPI) complex and ssDNA. They were devoid of polyreactivity. Nucleotide sequence analysis of the variable domain of the mu chain indicated the utilization of the VH4 71.2 gene or one allotypic variant, DXP4 and JH3 segments. The lambda light chain used the single gene from the V lambda 8 subfamily, J lambda 3 and C lambda 3 genes. The VH gene displayed 11 nucleotide changes in comparison with its putative germline counterpart. However, these nucleotide changes correspond to variations observed in other published VH4 sequences, suggesting gene polymorphism rather than somatic mutation. DXP4 and JH3 were also in germline configuration. The VL gene exhibited a single replacement mutation in CDR1. These data suggest that the monoclonal CLL B cells in this patient retained VH and VL genes in germline configuration although they secreted a pathogenic anti-cardiolipin antibody associated with clinical symptoms, vasculitis and thrombosis, which may be provoked by antibodies to the phospholipid/beta 2GPI complex.

Somatic mutation and memory

The germinal center plays a crucial role in the development of the memory B cell. The repertoire of the antigen-specific B cell is shaped in this microenvironment. Self-specific B cells escaping normal regulation may develop into high affinity pathogenic Ig-producing cells.