Early rearrangements of genes encoding murine immunoglobulin kappa chains, unlike genes encoding heavy chains, use variable gene segments dispersed throughout the locus

Multiple factors influence the contribution of individual immunoglobulin light chain genes to the naïve antibody repertoire

Background

The naïve antibody repertoire is initially dependent upon the number of germline V(D)J genes and the ability of recombined heavy and light chains to pair. Individual VH and VL genes are not equally represented in naïve mature B cells, suggesting that positive and negative selection also shape the antibody repertoire. Among the three member murine Vκ10 L chain family, the Vκ10C gene is under-represented in the antibody repertoire. Although it is structurally functional and accessible to both transcriptional and recombination machinery, the Vκ10C promoter is inefficient in pre-B cell lines and productive Vκ10C rearrangements are lost as development progresses from pre-B cells through mature B cells. This study examined VH/Vκ10 pairing, promoter mutations, Vκ10 transcript levels and receptor editing as possible factors that are responsible for loss of productive Vκ10C rearrangements in developing B cells.

Results

We demonstrate that the loss of Vκ10C expression is not due to an inability to pair with H chains, but is likely due to a combination of other factors. Levels of mRNA are low in sorted pre-B cells and undetectable in B cells. Mutation of a single base in the three prime region of the Vκ10C promoter increases Vκ10C promoter function in pre-B cell lines. Pre-B and B cells harbor disproportionate levels of receptor-edited productive Vκ10C rearrangements.

Conclusions

Our findings suggest that the weak Vκ10C promoter initially limits the amount of available Vκ10C L chain for pairing with H chains, resulting in sub-threshold levels of cell surface B cell receptors, insufficient tonic signaling and subsequent receptor editing to limit the numbers of Vκ10C-expressing B cells emigrating from the bone marrow to the periphery.

Skewed primary Igκ repertoire and V-J joining in C57BL/6 mice: implications for recombination accessibility and receptor editing

Previous estimates of the diversity of the mouse Ab repertoire have been based on fragmentary data as a result of many technical limitations, in particular, the many samples necessary to provide adequate coverage. In this study, we used 5'-coding end amplification of Igκ mRNAs from bone marrow, splenic, and lymph node B cells of C57BL/6 mice combined with amplicon pyrosequencing to assess the functional and nonfunctional Vκ repertoire. To evaluate the potential effects of receptor editing, we also compared V/J associations and usage in bone marrows of mouse mutants under constitutive negative selection or an altered ability to undergo secondary recombination. To focus on preimmune B cells, our cell sorting strategy excluded memory B cells and plasma cells. Analysis of ~90 Mbp, representing >250,000 individual transcripts from 59 mice, revealed that 101 distinct functional Vκ genes are used but at frequencies ranging from ~0.001 to ~10%. Usage of seven Vκ genes made up >40% of the repertoire. A small class of transcripts from apparently nonfunctional Vκ genes was found, as were occasional transcripts from several apparently functional genes that carry aberrant recombination signals. Of 404 potential V-J combinations (101 Vκs × 4 Jκs), 398 (98.5%) were found at least once in our sample. For most Vκ transcripts, all Jκs were used, but V-J association biases were common. Usage patterns were remarkably stable in different selective conditions. Overall, the primary κ repertoire is highly skewed by preferred rearrangements, limiting Ab diversity, but potentially facilitating receptor editing.

B7 and 34B7 Monoclonal Antibodies: A Theoretical Approach to the Molecular Basis of Immunoglobulin Cross-Reactive Antibodies

Neonatal natural antibodies (NAbs) are characterized by their high degree of idiotypic cross reactivity, together with some restrictions in the genetic mechanisms of variable region diversity. We report here the immunogenetic analysis of two anti-idiotype antibodies (B7 and 34B7 monoclonal antibodies [MAbs]), which are also polyreactive as NAbs. Evidence of a process of somatic mutations were found for heavy and light chain variable regions of both antibodies. A phylogenetic analysis of the V(H)J558 family showed that the immunoglobulin cross-reactivity displayed by B7 and 34B7 MAbs is not restricted to a particular subgroup of this family. Moreover, we identified amino acid motifs in the CDR H1 and H2 of B7 and 34B7 MAbs that are also present in high proportion in immunoglobulin cross-reactive antibodies (ICRA) reported in the Kabat database. We propose that these regions are involved in ICRA activity.

Analysis of B cell receptor production and rearrangement. Part I. Light chain rearrangement

A probabilistic model of allelic exclusion fails to explain the status of receptor genes and the receptor phenotype of most B cells. A large proportion of B cells have incompletely rearranged H and/or L chain genes (e.g. kappa0/kappa+) and most B cells express only one receptor. These properties seem to require deterministic features of B cell development such as special mechanisms that stop rearrangement. However, receptor editing has revealed that rearrangement-stop is not stable and that multi-receptor lymphocytes make up a significant fraction of certain B and T cell populations. Consequently we have revived the purely probabilistic approach in a model that now includes receptor editing and allows for some multi-receptor B cells. We find that this model can explain the observed properties of B cells when the frequency of self-reactive B cells is high. Indeed, as we illustrate for anti-DNA, this is the case. Hence the probabilistic model has life and assiduous use of the model suggests unexpected but not unrealistic features of lymphocyte development.

Editors and editing of anti-DNA receptors

Receptor editing is a means by which immature bone marrow B cells can become self-tolerant. Rearrangements of heavy (H) and/or light (L) chain genes are induced by encounter with autoantigens to change the specificity from self to nonself. We have developed site-directed transgenic mice (sd-tg) whose transgenes code for the H chain of antibodies that bind DNA. B cells that express the transgenic H chain associate mainly with four of the 93 functional Vkappa genes of the mouse. Numerous aspartate residues that might inhibit DNA binding by the V(H) domain distinguish these L chain Vkappa sequences, but engaging these Vkappa editors often requires multiple rearrangements. Among the edited B cells is a subset of multispecific cells that express multiple receptors. One consequence of multispecificity is partial autoreactivity; these multispecific B cells may contribute to autoimmunity.

Early Vlambda diversification in sheep

This study examined a number of tissues during early gestation in foetal sheep to determine the earliest site of Vlambda expression and time of generation of the Vlambda repertoire. Tissues, including spleen, liver, gut, blood and bone marrow, were obtained from 48, 55, 60 and 63 gestational day (g.d.) ovine foetuses and cDNA libraries were prepared from them by reverse transcription-polymerase chain reaction. Clones were randomly selected from cDNA libraries and subjected to sequencing. Analysis of these sequences and comparison with a pool of germline genes led to the following conclusions. The expression of Vlambda occurs earlier in spleen (48 g.d.) than in all of the other tissues examined. Also, diversity is seen earlier and at higher levels in early foetal spleen than in all of the other tissues examined. In this regard, it is notable that splenic Vlambda expression is readily apparent even before such gut-associated lymphoid tissue as the ileal Peyer's patch (IPP) has developed. Two germline Vlambda genes, 5.1 and 5.3 predominate in early immunoglobulin lambda light-chain gene rearrangement. Examination of Jlambda usage revealed the existence of a new Jlambda gene and its utilization during the early phases of the development of the ovine antibody repertoire. This study indicates that sites other than the IPP contribute to the diversification of the Vlambda repertoire in sheep. We suggest that it is likely that foetal spleen may provide a partially diversified B-cell repertoire before the IPP becomes active as a major site for massive clonal expansion and extensive diversification of B cells.

Innate antibody catalysis

Catalysis by antibodies is often assumed to require immunization with artificial haptens, which are proposed to stimulate adaptive immune processes and enable the development of catalytic sites with the ability to bind the transition state. Contrary to this assumption, we describe here a serine protease-like catalytic triad in an antibody light chain raised by immunization with vasoactive intestinal peptide (VIP), the structure and function of which is inherited via a germline V(L) gene. The serine protease mechanism was evident from loss of the catalytic activity by site directed mutagenesis at a framework region residue Asp1 (present study) and at two complementarity determining region residues Ser27a and His 93 (Gao, Q-S., Sun, M., Rees, A., Paul, S., 1995. Site-directed mutagenesis of proteolytic antibody light chain. J. Mol. Biol. 253, 658-664). All three catalytic residues (Ser27a, His93, Asp1) are also present in the germline counterpart of the mature V(L) gene, but the mature and germline sequences differ by four amino acids remote from the catalytic site. Reversion mutations were introduced at these amino acids in the mature light chain (His27 d:Asp, Thr28e:Ser, Ile34:Asn, Gln96:Trp; Kabat numbering, germline encoded residues shown second), generating the germline configuration of the protein. The germline light chain expressed peptidase activity, determined by assaying the cleavage of VIP and a synthetic protease substrate, Pro-Phe-Arg-Methylcoumarinamide. Differences between the kinetic constants for the mature and germline light chains were marginal. Diisopropylfluorophosphate, a serine protease inhibitor, blocked the peptidase activity of the germline light chain, suggesting the presence of the catalytic triad in a functional state. Like the mature light chain, the germline protein preferentially cleaves peptide bonds on the C-terminal side of basic residues. We conclude that the catalytic activity of certain antibodies is an innate function, originating over the course of phylogenetic evolution of the V(L) genes, as opposed to somatic processes.

Genetic origin of IgG antibodies cloned by phage display and anti-idiotypic panning from three patients with autoimmune thrombocytopenia

The beneficial use of intravenous immunoglobulins (IVIG) in certain groups of patients with autoimmune thrombocytopenic purpura (AITP) has been proven. AITP is a severe disease in children with a still unknown etiology. It is not clear how IVIG functions in this and other autoimmune diseases. To analyze and compare patient-derived monoclonal IgG antibodies that are bound by IVIG in an anti-idiotypic manner, the combinatorial antibody phage display system was applied. From three different patients with AITP, a large number of clones specifically reacting with IVIG molecules were enriched. The heavy and light chain variable regions were sequenced and compared with each other and with databases. Many variable regions showed extensive replacement mutations within the complementarity-determining regions, while two were identical to germ-line genes. Our data show that the most frequently used germ-line gene loci of these IVIG binders are identical to those observed for many other autoantibodies. This implicates a specific interaction of IVIG particularly with autoantibodies and B cell receptors derived from germ-line genes that are often used for the generation of autoantibodies.

Asymmetric Contribution to Ig Repertoire Diversity by Vκ Exons: Differences in the Utilization of Vκ10 Exons

The mouse has approximately 140 germline Vκ genes, and functional Vκ exons are expressed at roughly equivalent levels in the preimmune repertoire. We have examined the expression of individual members of the Vκ10 family. Vκ10A and Vκ10B genes have been utilized in numerous hybridomas and myelomas, while Vκ10C has not. In this study, we have cloned the Vκ10C gene and shown that it is structurally functional, has the expected promoter elements and recombination signal sequences, and that it is capable of recombination. Vκ10C mRNA, however, is present at levels at least 1000-fold lower than Vκ10A and Vκ10B in adult spleens. While there are no sequence differences in the octamer or TATA box between Vκ10C and Vκ10A, there are three nucleotide changes in the promoter region. These promoters equally drive the expression of a reporter gene in B cells or plasma cells, but the Vκ10A promoter is able to drive expression in pre-B cell lines significantly better than the Vκ10C promoter (p < 0.05). Vκ10C rearrangements can be detected in bone marrow and splenic DNA. Therefore, the lack of Vκ10C expression may reflect the inability of Vκ10C-rearranged cells to undergo positive or negative selection. Our results suggest that the available Ab repertoire is shaped not only by the number of structurally functional genes, but also by the ability of assembled genes to be expressed at critical points during B cell maturation.

Molecular Analysis of Polyreactive Monoclonal Antibodies from Rheumatic Carditis: Human Anti-N-Acetylglucosamine/Anti-Myosin Antibody V Region Genes

Anti-myosin Abs are associated with inflammatory heart diseases such as rheumatic carditis and myocarditis. In this study, human cross-reactive anti-streptococcal/anti-myosin mAbs 1.C8, 1.H9, 5.G3, and 3.B6, produced from peripheral blood lymphocytes of patients with rheumatic carditis, and mAb 10.2.5, produced from a tonsil, were characterized, and the nucleotide sequences of their VH and VL genes were analyzed. Human mAbs 1.C8, 1.H9, 10.2.5, and 3.B6 reacted with human cardiac myosin while mAb 5.G3 did not. The mAbs were strongly reactive with N-acetyl-β-d-glucosamine, the dominant epitope of the group A streptococcal carbohydrate. mAb 1.H9 was moderately cytotoxic to rat heart cells in vitro in the presence of complement. The anti-myosin mAbs from rheumatic carditis were found to react with specific peptides from the light meromyosin region of the human cardiac myosin molecule. Anti-streptococcal/anti-myosin mAbs from normal individuals reacted with distinctly different light meromyosin peptides. The mAbs were encoded by VH3 gene segments V3-8, V3-23, and V3-30 and by the VH4 gene segment V4-59. The variable region genes encoding the anti-streptococcal/anti-myosin repertoire were heterogeneous and exhibited little evidence of Ag-driven somatic mutation.

Rearrangement and Selection in the Developing Vκ Repertoire of the Mouse: An Analysis of the Usage of Two Vκ Gene Segments

Detailed analysis of the rearrangement and expression of two mouse Vκ genes has been used to examine B cell repertoire development. The Vκ1-A gene is used by a large proportion (9.6%) of splenic B cells in the adult primary repertoire, whereas the Vκ22 gene is used at a much lower frequency (0.16%). Consistent with these results, quantitative PCR (Q-PCR) assays revealed that the number of splenic B cells with rearranged Vκ1-A genes is much greater than the number with rearranged Vκ22 genes. Q-PCR was also performed on both normal bone marrow pre-B cells and transformed pre-B cells induced to rearrange their κ loci at high frequency. In contrast to splenic B cell rearrangements, the numbers of Vκ1-A and Vκ22 rearrangements in pre-B cells differ by only two- or threefold, suggesting that the intrinsic rearrangement frequencies of these two Vκ genes are not significantly different. Further evidence of disproportionate selection was obtained by comparing the percentages of productive rearrangements amplified from genomic splenic DNA. Sequence analysis showed 84% (37 of 44) of the Vκ1-A rearrangements but only 57% (29 of 51) of the Vκ22 rearrangements to be in-frame. Together these results suggest that B cells expressing Vκ1-A-encoded light chains are preferentially selected either in the periphery or in the transition from pre-B to B cell. Sequence data also reveal a surprisingly restricted diversity of VJ junctions, apparently due to biases introduced by the rearrangement mechanism.

The 3' part of the immunoglobulin kappa locus of the mouse

A detailed restriction map of a 430-kb contig comprising the single Ckappa, the 5 Jkappa and the adjoining 22 Vkappa gene segments is presented. The first 12 Vkappa genes following the JkappaCkappa region belong to the Vkappa21 family, the subsequent ones to the closely related families Vkappa8 and Vkappal 9/ 28. Previous difficulties in cloning all Vkappa21 genes can now be explained by the presence of a duplicated region in this part of the locus. The structure was established by analysis of yeast artificial chromosome, bacterial artificial chromosome and cosmid clones and by the so-called long template PCR technique. The distance between Ckappa and the proximal Vkappa21 gene is 22 kb and the average distances between the Vkappa genes are about 20 kb. Of the 12 Vkappa21 genes 5 were sequenced for the first time and 8 of the 12 genes were found to be expressed. Of the 10 Vkappa8 and Vkappa19/28 germline genes 9 are new; expression products of 8 of the 10 genes were known. The known 5', 3' polarities allow to specify for the 22 Vkappa genes whether they are rearranged to the JkappaCkappa element by a deletion or an inversion mechanism. Also the formation of interesting rearrangement products in classical cell lines as MPC11, MOPC41 and PC 7043 can be explained now. The non-Vkappa sequence L10 whose rearrangement by inversion has been described earlier (Hoechtl and Zachau, Nature 1983. 302: 260-263) was now localized downstream of JkappaCkappa.

A recombinant single chain antibody neutralizes coronavirus infectivity but only slightly delays lethal infection of mice

The variable region genes of a murine anti-coronavirus monoclonal antibody (mAb) were joined by assembly polymerase chain reaction and expressed in Escherichia coli in a single chain variable fragment (scFv) configuration. After induction of expression, the expected 32-kDa protein was identified by Western immunoblotting with specific rabbit anti-idiotype antibodies. The scFv fragments were purified from soluble cytoplasmic preparations by affinity chromatography on nickel agarose, which was possible with an N-terminal but not with a C-terminal histidine tag. Purified scFv fragments retained the antigen-binding properties of the parental antibody, could inhibit its binding to viral antigens with apparently higher efficiency than monovalent antigen-binding (Fab) fragments, but neutralized viral infectivity with lower efficiency (about sevenfold at a molar level). To evaluate the usefulness of these smaller and less immunogenic molecules in the treatment of viral diseases, mice were treated with purified recombinant scFv fragments and challenged with a lethal viral dose. A small delay in mortality was observed for the scFv-treated animals. Therefore, even though the scFv could neutralize viral infectivity in vitro, the same quantity of fragments that partially protected mice in the form of Fab only slightly delayed virus-induced lethality when injected as scFv fragments, probably because of a much faster in vivo clearance: the biologic half-life was estimated to be about 6 min. Since a scFv derived from a highly neutralizing and protective mAb is only marginally effective in the passive protection of mice from lethal viral infection, the use of such reagents for viral immunotherapy will require strategies to overcome stability limitations.

Molecular cloning, expression, and characterization of a functional single-chain Fv antibody to the mycotoxin zearalenone

The heavy-chain and kappa light-chain variable region genes of an antizearalenone hybridoma cell line (2G3-6E3-2E2) were isolated by PCR and joined by a DNA linker encoding peptide (Gly4Ser)3 as a single-chain Fv (scFv) DNA fragment. The scFv DNA fragment was cloned into a phagemid (pCANTAB5E) and expressed as a fusion protein with E tag and phage M13 p3 in Escherichia coli TG1. In the presence of helper phage M13K07, the scFv fusion protein was displayed on the surfaces of recombinant phages. High-affinity scFv phages were enriched through affinity selection in microtiter wells coated with zearalenone-ovalbumin conjugate. The selected recombinant phages were used to infect E. coli HB2151 for the production of soluble scFv antibodies. One selected clone (pQY1.5) in HB2151 secreted a soluble scFv antibody (QY1.5) with a high zearalenone-binding affinity (concentration required for 50% inhibition of binding, 14 ng/ml), similar to that of parent monoclonal antibody in a competitive indirect enzyme-linked immunosorbent assay. However, scFv QY1.5 exhibited higher cross-reactivity with zearalenone analogs and had greater sensitivity to methanol destabilization than the parent monoclonal antibody did. Nucleotide sequence analyses revealed that the light-chain portion of scFv QY1.5 had a nucleotide sequence identity of 97% to a mouse germ line gene VK23.32 in mouse kappa light-chain variable region subgroup V, whereas the heavy-chain nucleotide sequence was classified as mouse heavy-chain subgroup III (D) but without any closely related members having highly homologous complementarity-determining region sequences. The potential of soluble scFv QY1.5 for routine screening of zearalenone and its analogs was demonstrated with zearalenone-spiked corn extracts.

Immunoglobulin gene diversification in cattle

Research in several species has revealed that different types of mammals have evolved divergent molecular and cellular strategies for generating immunoglobulin diversity. Other chapters in this text have highlighted the specific characteristics unique to chicken, rabbit, mouse, human and sheep B lymphocyte development; namely indicating differences in the mechanisms of diversity and the site of primary B cell development. Studies of the bovine system have indicated that like the sheep system, the ileal Peyer's patch (IPP) is a likely chicken bursal equivalent, and is a site of primary B lymphocyte development. Substantial investigation in sheep has indicated that Ig diversity is created by untemplated somatic mutation and intense selective pressure (Reynaud et al., 1991). The frequency of alteration in the sheep Ig light chain gene locus also is characteristic of the bovine system, however, recent evidence from sequencing of bovine lambda light chain genes indicates that one mechanism that contributes to diversity is gene conversion, utilizing several pseudogenes located in the Ig locus (Parng et al., 1996). The mechanism by which this hyperalteration of Ig genes occurs in both sheep and cattle is poorly understood and is thus the focus of considerable investigation. The study of events in the IPP may also have informative ramifications for secondary diversification of the Ig repertoire by somatic hyperalteration in germinal centers.

Various V-J rearrangement efficiencies shape the mouse lambda B cell repertoire

The diversity of the B cell repertoire of C kappa knockout mice is limited by the expression of four lambda light chain types. Among the spleen B cells, lambda 1 is expressed by the majority (58%) of cells, and lambda 3 by the minority (8%), while lambda 2 (V2) and lambda 2 (Vx) are expressed in intermediate quantities (18% and 16%, respectively). To assess the influence of mechanistic pressures on the lambda subtype distribution, the proportions of the different lambda rearrangements were determined in various B cell subpopulations divided on the basis of the lambda subtype expressed, and the V lambda J lambda junction sequences were studied at different steps of B cell differentiation (pre-B, immature and mature B cells). The data show that (1) the ratio of productive/non-productive VJ junctions is determined by the nature of the lambda segments that are rearranged as can be observed in the pre-B cells, (2) V1-J1 non-productive rearrangements are often found in the lambda 1-negative B cells in the periphery, and (3) V1J3 junctions are often non-productive regardless of the nature of the cells analyzed. Our results, therefore, suggest that a strong probability of initiating a V1-J1 rearrangement and a weak probability of giving a productive V1J3 junction are responsible for the lambda 1 dominance and the lambda 3 under-expression, respectively. The intermediate proportion of lambda 2(V2) subtype is most likely due to a probability of obtaining a productive joint that is better than that for V1J3 and a probability of initiating a rearrangement that is lower than that for V1J1. However, the lambda 2(Vx) cell proportion cannot be determined only by these parameters.

Idiotypic diversity and variable region gene usage by mouse anti-HLA-DQ3 mAb

The anti-HLA-DQ3 monoclonal antibodies (mAb) KS13, SO1, SO2, SO3, SO4, and SO5 recognize spatially close but distinct antigenic determinants, since they crossinhibit each other in their binding to HLA-DQ3 antigens, but do not share idiotopes recognized in their antigen combining site by syngeneic and anti-id antisera and mAb. Furthermore, mAb SO1, SO3, SO4, and SO5 react also with HLA-DQ allospecificities other than HLA-DQ3. Sequence analysis of the heavy (VH) and light (VL) chain variable region of the six mAb revealed preferential usage of VH 36-60 and VK 12/13 gene families. However, the individual VH and VL germline gene usage by the six mAb is diverse and the utilization of D, JH, and JL gene segments is heterogeneous. The diverse usage of VH and VL gene segments and heterogeneous amino acid sequences of VH and VL CDR, together with the heterogeneous idiotypic profile, may reflect the complexity of the determinants recognized by the six mAb on HLA-DQ3 antigens. The results we have presented provide for the first time information about the structural basis of the diversity of antibodies recognizing human histocompatibility antigens.

The nude mutation results in impaired primary antibody repertoire

We have studied the effect of the nude mutation and/or T lymphocytes on the development of V gene germ-line repertoire in neonatal athymic (nu/nu) and euthymic (+/nu) littermates. A total of 2.35 x 10(6) and 1.47 x 10(6) B lymphocyte clones from nu/nu and +/nu neonates, respectively, were examined for the expression of select VH (J558, J606, S107, 36-60, 7183 and Q52) and Vx (1, 2, 8 and 9) gene families as well as VH (J558, S107) + Vx (1, 9) associations. Data showed that the nude mutation, whether homozygous or heterozygous, significantly affects VH and Vx gene expression as well as VH and Vx pairings and, thus, provide evidence for a defective development of B cell repertoire in both athymic nude (nu/nu) and euthymic (+/nu) mice. In addition, an analysis of 3.34 x 10(6) B lymphocyte clones from adult C57BL/6 mice showed non-stochastic association between VHJ558 + Vx1 gene families that suggests restrictions on clonal population in order to maintain homeostasis in the immune system. Studies outlined here, therefore, describe an hitherto unknown defect in the development of B lymphocyte repertoire as a result of the nude mutation which is independent of thymic dysgenesis.

Immunoglobulin light chain class multiplicity and alternative organizational forms in early vertebrate phylogeny

The prototypic chondrichthyan immunoglobulin (Ig) light chain type (type I) isolated from Heterodontus francisci (horned shark) has a clustered organization in which variable (V), joining (J), and constant (C) elements are in relatively close linkage (V-J-C). Using a polymerase chain reaction-based approach on a light chain peptide sequence from the holocephalan, Hydrolagus colliei (spotted ratfish), it was possible to isolate members of a second light chain gene family. A probe to this light chain (type II) detects homologs in two orders of elasmobranchs, Heterodontus, a galeomorph and Raja erinacea (little skate), a batoid, suggesting that this light chain type may be present throughout the cartilaginous fishes. In all cases, V, J, and C regions of the type II gene are arranged in closely linked clusters typical of all known Ig genes in cartilaginous fishes. All representatives of this type II gene family are joined in the germline. A third (kappa-like) light chain type from Heterodontus is described. These findings establish that a degree of light chain class complexity comparable to that of the mammals is present in the most phylogenetically distant extant jawed vertebrates and that the phenomenon of germline-joined (pre-rearranged) genes, described originally in the heavy chain genes of cartilaginous fishes, extends to light chain genes.

Variable region gene selection of immunoglobulin G-expressing B cells with specificity for a defined epitope on type II collagen

Immunization with type II collagen (CII) induces collagen-induced arthritis (CIA) in animals, and B cells reactive with CII are involved in the induction and manifestation of the disease. In this study, B cell hybridomas producing IgG antibodies specific for a major epitope on mouse CII (the "C1" epitope, amino acid 316-333), were isolated 11 days after immunization from draining lymph nodes in DBA/1 mice. Injection into neonatal mice of purified and biotinylated monoclonal antibodies binding the C1 epitope led to a specific binding to joint cartilage, demonstrating that the antibodies interact with native antigen in vivo. cDNA sequencing of the B cell clones revealed that they all expressed the same combination of a variable heavy chain (VH J558 family) and light chain (V kappa 21 family) germ-line gene, apparently lacking somatic mutations. The presence of isotype-switched B cells expressing a certain combination of V genes encoding antibodies that bind epitopes in vivo, indicates that this B cell population has been peripherally selected.

Restricted kappa chain expression in early ontogeny: biased utilization of V kappa exons and preferential V kappa-J kappa recombinations

To determine the extent of kappa chain diversity in the preimmune repertoire early in development, kappa cDNA libraries were analyzed from 15-d old fetal omentum, 18-d-old fetal liver, and 3-wk old bone marrow. An anchored polymerase chain reaction approach was used to avoid bias for particular V kappa families. From the sequence analysis of 27 bone marrow clones, 10 different families and 20 unique V kappa genes were identified. In contrast, the V kappa expression in the fetus is highly restricted and clearly differs from the broader distribution see in 3-wk-old bone marrow. Although several V kappa families were represented in the fetal library including V kappa 9, V kappa 10, V kappa 4,5, V kappa 8, and V kappa 1, one or two members of individual families were observed repeatedly. The fetal liver and omentum libraries were found to be largely overlapping. Given the V kappa families/exons identified in the fetal sequences, the mechanism of kappa rearrangements in the early repertoire appears to occur predominantly by inversion. Importantly, the fetal repertoire was further restricted by dominant V kappa-J kappa combinations such as V kappa 4,5-J kappa 5, V kappa 9-J kappa 4, and V kappa 10-J kappa 1. Since in some cases independent rearrangements could be established, the results indicate a bias for particular V kappa-J kappa joins. The results also suggest that clonal expansion/selection in the fetal repertoire takes place after light chain rearrangement as opposed to at the pre-B cell level in the bone marrow. The restriction observed in kappa light chain expression together with known restrictions in gene usage and junctional diversity at the heavy chain level indicate a remarkably conserved fetal repertoire.

V kappa gene segments rearranged in chronic lymphocytic leukemia are distributed over a large portion of the V kappa locus and do not show somatic mutation

The structure of the human V kappa locus has been thoroughly investigated, but how the germ-line V kappa gene segment repertoire is actually sampled in kappa chain gene rearrangements is not known. In order to begin to answer this question we have polymerase chain reaction (PCR) amplified the rearranged V kappa genes from 26 kappa-expressing cases of chronic lymphocytic leukemia (CLL), followed by cloning and sequencing of the PCR product. All four V kappa gene families were represented amongst rearranged genes. In 25 out of 32 cases, the sequence of the rearranged gene matches perfectly that of 1 of 11 different known germ-line V kappa genes, indicating that no somatic mutation has occurred. Of the remaining 7 rearranged V kappa genes, 4 differ from known germ-line genes by only one or two amino acid residues; and 3 differ from each other and from all known sequences by 5 or more residues, suggesting that somatic mutation has occurred in these 3 cases. We conclude that: (a) in at least three-quarters of cases the rearranged genes are unmutated; (b) there is preferential usage of individual V kappa genes but not of V kappa gene families; and (c) the V kappa genes used are widely dispersed in the V kappa locus.

Molecular analysis of anti-idiotypic monoclonal antibodies in the HLA-DR antigenic system

The structural organization of anti-idiotypic (id) antibodies has been investigated mostly in haptenic systems. No information is available about the structural characteristics of anti-id antibodies in major histocompatibility complex (MHC) antigenic systems, although these data may contribute to our understanding of the molecular basis of their functional role in the immune response. Therefore, we have determined the nucleotide and derived amino acid sequence of the VH and VL regions of the anti-id monoclonal antibodies (mAb) F5-444, F5-830, F5-963, F5-1126, F5-1336 and F5-1419, which had been elicited with the syngeneic anti-HLA-DR1, 4, w14, w8, 9 mAb AC1.59. The six anti-id mAb are heterogenous in their VH and VL region gene usage. This structural heterogeneity is not correlated with their target specificity and with their ability to elicit anti-HLA-DR antibodies. The latter characteristic is markedly influenced by a limited number of amino acid substitutions, since mAb F5-444, which induces anti-HLA-DR antibodies, differs only in two residues in complementarity-determining regions and in five residues in framework regions from mAb F5-1126, which does not induce anti-HLA-DR antibodies. The heterogeneity in VH and VL region gene usage by the six anti-id mAb in the HLA-DR system is at variance with the restricted VH and VL region gene usage by syngeneic anti-id mAb in several haptenic systems. Furthermore, at variance with haptenic systems, the primary structure of the D segments of the anti-id mAb is not correlated with their ability to induce anti-HLA-DR antibodies. On the other hand, the frequency of D-D fusion events underlying the derivation of the D segments of the six anti-id mAb in the HLA-DR system and their average length are similar to those found in anti-id mAb in haptenic systems. In addition, like in the latter systems, somatic mutations appear to contribute to the generation of diversity of anti-id mAb in the HLA-DR system.

Both IgM and IgG anti-DNA antibodies are the products of clonally selective B cell stimulation in (NZB x NZW)F1 mice

Disease activity in systemic lupus erythematosus is closely associated with the appearance of immunoglobulin (Ig)G antibody to native DNA in both humans and mice. Like normal antibody responses, the anti-DNA autoantibody first appears as IgM and then switches to IgG. Structural studies of IgG anti-DNA suggest that these antibodies are the products of clonally selected, specifically stimulated B cells. The origins of the IgM anti-DNA have been less clear. To determine whether the earlier appearing IgM anti-DNA antibody in autoimmune mice also derives from clonally selected, specifically stimulated B cells or B cells activated by nonselective, polyclonal stimuli, we have analyzed the molecular and serological characteristics of a large number of monoclonal IgM anti-DNA antibodies from autoimmune (NZB x NZW)F1 mice. We have also analyzed IgM and IgG anti-DNA hybridomas obtained from the same individual mice to determine how the later-appearing IgG autoantibody may be related to the earlier-appearing IgM autoantibody within an individual mouse. The results demonstrate that: (a) IgM anti-DNA, like IgG, has the characteristics of a specifically stimulated antibody; (b) IgM and IgG anti-DNA antibodies have similar variable region structures and within individual mice may be produced by B cells derived from the same clonal precursors; (c) recurrent germline and somatically derived VH and VL structures may influence the specificity of anti-DNA monoclonal antibody for denatured vs. native DNA; and (d) the results provide a structural explanation for the selective development of IgG antibody to native DNA as autoimmunity to DNA progresses in (NZB x NZW)F1 mice.

The fetal omentum in mice and humans. A site enriched for precursors of CD5 B cells early in development

From these studies the fetal omentum appears to be an important site of B-cell generation in humans and a source of CD5 B cells in mice. We have analyzed the fetal omentum in other species and have found that B-cell development as determined by the presence of cytoplasmic IgM+ pre-B cells is also detected in the fetal rabbit omentum. We do not know if there is bias towards the production of CD5 B cells in this species; however, these preliminary results demonstrate that this site may be conserved throughout evolution in mammals as a site of B-cell generation. Because the fetal liver is also a source of precursors that can reconstitute this B-cell subset, what is the relationship between omentum and liver during the development of Ly-1 B cells? The most obvious relationship between these two sites is that cells simply migrate from one location to the other; that is, precursor cells may migrate from the fetal liver into the fetal omentum and in this milieu give rise to exclusively Ly-1+ B cells or the sister population. Alternatively, precursors of Ly-1 B cells may arise in the omentum and migrate to the liver. This is demonstrated graphically in the diagram (Fig. 6a) of a transverse section through an 8-week human fetus. In this paper, however, we suggest a model for the development of Ly-1+ B cells from the omentum and liver in which Ly-1 B cells arise from distinct precursors located in situ in the mesodermally derived omentum and mesothelial-derived liver capsule. The omentum primordia forms as the back to back fusion of the mesodermally derived lining of the peritoneal cavity, and this lining surrounds the developing gut when the liver begins to develop as an outgrowth of the intestinal primordia at approximately 3.5 weeks gestation; the outer covering or capsule of the liver is derived from the same tissue of origin as the omentum. Figure 6B is a diagram of a section through the same plane as Figure 6A but the body wall has been omitted. We propose that the Ly-1+ B cells arise in situ in the omentum and lining of the liver as indicated in Figure 6B. That Ly-1+ B cells arise from distinct precursors has been suggested by others, but ours is the first evidence for a developmental site that apparently contains B-cell progenitors for this B-cell subset.

Interrelating B cell subpopulations and environmental regulation with the expression of three tiers of repertoire diversity

The B cell repertoire consists of three tiers of clonotype diversity. One tier, which is the product of H chain V region rearrangements in the absence of N additions, is of limited diversity (less than 10(8) clonotypes) so that clonotypes of this tier would be expected to recur within and among B cells of individuals of an inbred strain. These clonotypes, therefore, could be subjected to, and conserved by, evolutionary selective pressures such as those imposed by ubiquitous bacterial pathogens. The second tier of clonotypes is created by H chain V region rearrangements that include N additions, and is, therefore, exceedingly diverse. Clonotypes of this tier would be unlikely to recur; however, by providing maximal diversity they would ensure protection against a wide spectrum of pathogens. The third tier of diversity is that which is generated by the superimposition of somatic mutations on clonotypes of the other two tiers. This tier of clonotypes is reflective of the refinement of specificities that are destined for expression in memory B cells. B cells exists as three distinct subpopulations, Ly-1 B cells, conventional primary B cells and memory B cells. These subpopulations differ functionally, developmentally, and by the extent to which they are impacted by immunoregulatory processes. Furthermore, B cells of these subpopulations differentially express the three tiers of clonotype diversity.

Comparative expression of adult and fetal V gene repertoires

A hallmark of the immune system is the extraordinary diversity associated with antibodies. This is made possible by a series of genetic rearrangements involving variable region gene segments. Considerable detail is known about these genetic mechanisms except for the enzymatic machinery involved. An important question in studies of the generation of diversity is whether V genes are selected for rearrangement mainly in a random manner or selected by particular developmental rules. Past studies have indicated that the acquisition of fetal and neonatal specificity repertoires is a nonrandom process. In this report, we review our studies that directly compare the adult and fetal/neonatal V gene repertoires. The evidence suggests that the adult repertoire is more diverse with indications of a random use of VH gene families. However, whether V genes are indeed randomly used in the adult remains to be clarified at the VH gene member level. The fetal repertoire, on the other hand, appears nonrandom in V gene usage. In addition, the fetal repertoire is mostly germline encoded with little evidence of junctional diversity. Taken together, the results indicate different rules for generation of the adult and fetal repertoires, findings most likely explain by distinct B cell subsets and B cell progenitors at early stages in ontogeny.

Development and function of the early B cell repertoire

The early B cell repertoire is characterized by extensive interconnectivity, autoreactivity and multispecificity. Our preliminary sequence analysis of some of the idiotype specific antibodies is beginning to provide molecular clues to explain the observed multireactivity and the expression of shared idiotypic determinants on immunoglobulins of early B cells. The VH gene rearrangements analyzed are typical of the early pre-B cell and CD5 B cell repertoire. Some of these include shared or identical CDR3 regions resulting from the use of germline VH, D and JH gene segments in the absence of N region addition. As previously described, the most D proximal VH genes are also used most frequently. Collectively these genetic restrictions, together with the lack of somatic mutation, suggest that the characteristic self reactivity of the early B cell repertoire is related to the expression of germline gene segments and limited use of diversification mechanisms. It has also been possible for the first time to isolate hybridomas secreting functional IgM molecules which use the most D proximal VH gene, VH81X. These antibodies and another example from the VH7183 family have a broad multireactivity pattern possibly because of the presence of an unusually high number of charged amino acid groups present in the VH region. These findings are preliminary and more extensive studies are needed to establish if these groups are responsible for the highly cross-reactive nature of these antibodies. Nevertheless, these unusual characteristics signify a unique role for antibodies expressing this VH gene during B cell development. It is also clear that the observed anti-lymphocyte reactivity, another feature of the newborn repertoire, is the result of the prevalence of B cells using similar if not identical VHDJH genes and DJH joins. The development of these B cells appears to occur consistently in early ontogeny and, again, are not found in conventional splenic B cells obtained from the normal adult. Understanding the functional significance of the early appearance of these antibodies may help to clarify and understand their role during development as well as in autoimmunity. We propose that the unique self reactive nature of the early repertoire provides a pattern within which self-assertiveness develops and results in the establishment of the adult repertoire. In doing so, dominant clones are established which may or may not be within, but whose selection and differentiation is directed by the CD5 B cell subset.(ABSTRACT TRUNCATED AT 400 WORDS)

Two closely related kappa variable region pseudogenes pose an evolutionary paradox

Two pseudogenes belonging to the Igk-V1 variable region group have been isolated from BALB/c mice. The genes share greater than 96.5% identity of nucleotide sequence in a 1800 base pair (bp) region surrounding the coding region, but deletions of 221 bp and 84 bp have removed essential sequences from the two genes. As the deletions are different in the two pseudogenes, they must have occurred independently in each gene during or subsequent to the duplication event which gave rise to the genes from a common ancestral gene. Polymerase chain reaction analysis was used to identify the pseudogenes in inbred strains of mice. BALB/c (Igkc) and AKR (Igka), prototype strains representative of the predominant kappa haplotypes, possess both pseudogenes but no intact copy. Only one of the pseudogenes was present in SJL (Igka). Strains C58, c.C58 (Igkd) and NZB (Igkb) possessed an intact version of the gene. This distribution of haplotypes is consistent with a close linkage of the pseudogenes with other Igk-V1 genes on chromosome 6. The translated amino acid sequence of the pseudogenes indicates that prior to their acquiring deletions they encoded typical Igk-V1 variable regions except for an unusual FR2 region, in which the conserved proline at position 44 is replaced by leucine and the normally hydrophobic position 36 was occupied by histidine. Possible mechanisms to explain the occurrence of deletions in both of the pseudogenes in the recent evolution of BALB/c are discussed. One explanation would be that the two genes were already nonfunctional at the time of the duplication so that the subsequent deletions represent neutral events which became fixed in the inbred strains by a process of genetic drift. Alternatively, if the genes were functional at the time of duplication, their rapid loss due to deletion events suggests that negative selection may have acted to eliminate the genes from the V-region repertoire.

In situ studies of the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl. I. The architecture and dynamics of responding cell populations

After primary immunization with an immunogenic conjugate of (4-hydroxy-3-nitrophenyl)acetyl, two anatomically and phenotypically distinct populations of antibody-forming cells arise in the spleen. As early as 2 d after immunization, foci of antigen-binding B cells are observed along the periphery of the periarteriolar lymphoid sheaths. These foci expand, occupying as much as 1% of the splenic volume by day 8 of the response. Later, foci grow smaller and are virtually absent from the spleen by day 14. A second responding population, germinal center B cells, appear on day 8-10 and persist at least until day 16 post-immunization. Individual foci and germinal centers represent discrete pauciclonal populations that apparently undergo somatic evolution in the course of the primary response. We suggest that foci may represent regions of predominantly interclonal competition for antigen among unmutated B cells, while germinal centers are sites of intraclonal clonal competition between mutated sister lymphocytes.

Deletions of immunoglobulin C kappa region characterized by the circular excision products in mouse splenocytes

We have identified circular DNAs containing the kappa light chain constant region (C kappa), as well as the excision products of V kappa-J kappa and V lambda-J lambda joining in adult mouse splenocytes. Analysis of C kappa-positive circular DNA clones revealed two recombination sites (intron recombining sequence [IRS]1 and -2) within the germline J kappa-C kappa intron region and the recombining sequence (RS) located downstream of the C kappa exon. While RS contains a conserved heptamer and nonamer separated by a 23-bp spacer on the 5' side, IRS1 sequence is an isolated heptamer without an obvious nonamer, and IRS2 contains a variant heptamer, CACAAAA. Since IRS1 and IRS2 recombined with both RS (23-bp spacer signal) and V kappa (12-bp spacer signal) with significant frequency, intron recombination sites seem to have dual recombination signals. These findings provide direct evidence that C kappa deletion preceding lambda gene rearrangement can occur by looping out and excision. Increased accessibility of inefficient recombinational loci within the intron may enable recombinase to accept wide signal sequence variation.

V lambda-J lambda rearrangements are restricted within a V-J-C recombination unit in the mouse

The murine lambda gene locus is organized as follows: V lambda 2-V lambda x-J lambda 2C lambda 2-psi J lambda 4C lambda 4-V lambda 1-J lambda 3C lambda 3-J lambda 1C lambda 1 where all segments have the same transcriptional orientation. The combinatorial process of gene recombination should allow the generation of eight distinct immunoglobulin light chains. We have therefore investigated the probability of obtaining such chains among the mature lambda B cell repertoire. We analyze serum lambda immunoglobulins and lambda B cell clones induced by treatment with rabbit anti-lambda antibodies coupled to LPS. Confirming previous data obtained by others, our results indicate that the rearrangements of lambda segments take place within each V lambda-J lambda-C lambda cluster, thereby defining a unit of recombination. Our results also provide no evidence for the use of undescribed segments as has been recently suggested by the finding of the V lambda x segment.

Contribution of the VH11 gene family to mitogen-responsive B cell repertoire in C57BL/6 mice

The contribution of VH11 gene family to the development of the primary B cell repertoire has been studied by analyzing 1.8 x 10(4) mitogen induced B lymphocyte colonies. The data demonstrate that VH11 family is predominantly expressed among neonatal splenic as well as adult peritoneal B cell colonies, both rich in Ly-1+ B cells. VH11 gene family expression among B splenocytes decreases during ontogeny and VH11 family pairs stochastically with different V kappa families among mitogen-activated neonatal B cell colonies, which are representative of an antigen unselected B cell repertoire. Thus, an increased VH11 expression among peritoneal and neonatal B cells points towards its biased expression among Ly-1+ B lymphocytes. The restricted V gene rearrangements and VH11-V kappa 9 pairing observed among anti-bromelain-treated mouse red blood cells autoantibodies are likely to be an outcome of both intrinsic gene recombination processes per se as well as selection by an autoantigen and/or local selective environmental factors.

Lack of feedback inhibition of V kappa gene rearrangement by productively rearranged alleles

Circular DNAs excised by immunoglobulin kappa chain gene rearrangements were cloned and characterized. 16 of 17 clones examined were double recombination products containing a V kappa-J kappa rearrangement (coding joint) as well as the reciprocal element (signal joint) of another V kappa-J kappa rearrangement. These products suggested multiple recombination, primary inversion, and secondary excision. In primary events, 5 of 16 translational reading frames were in-phase. Thus, V kappa gene rearrangement may not be inhibited by the presence of a productively rearranged allele. An unusually large trinucleotide (P) insertion forming a palindrome of 12 nucleotides was also observed in one of the coding joints.

Evolutionary relationship between human and mouse immunoglobulin kappa light chain variable region genes

Similar to the Igh-V multigene family, the human or mouse Igk-V repertoire is a distorted continuum of homologous genes that may be grouped into families displaying greater than 80% nucleic acid sequence similarity among their members. Systematic interspecies sequence comparisons reveal that most human Igk-V gene families exhibit clear homology to mouse Igk-V families (sequence similarity generally greater than 74%). A hypothetical phylogenetic tree of Igk-V genes predicts that a minimum of seven Igk-V genes/families predate mammalian radiation. In two cases, several interrelated mouse Igk-V families exhibit phylogenetic equidistance with just one human Igk-V family, implying a more pronounced divergence for the elevated number of Igk-V gene families in the mouse. Mouse-human Igk-V comparisons, moreover, illustrate how expansion, contraction, and perhaps deletion of Igk-V gene families shape the Igk-V repertoire during mammalian evolution.

A single pre-B cell can give rise to antigen-specific B cells that utilize distinct immunoglobulin gene rearrangements

A group of hybridomas that express antibodies with related specificities for the influenza virus hemagglutinin (HA), that represent B cells that were the clonal progeny of a single pre-B cell, and that utilized distinct L chain gene rearrangements have been characterized. The clonal relationship was established by the sharing of H chain gene rearrangements at both the productive and the nonproductive alleles. Among these hybridomas, one group had rearranged only one of its kappa alleles, having joined a V kappa 24 gene to the J kappa 2 gene segment. The other group utilized the same V kappa 24 gene segment in productive rearrangement to the J kappa 5 gene segment, and shared an aberrant rearrangements among members of the same B cell clone can normally occur, and can contribute to the generation and diversification of the immune repertoire that is available for the recognition of foreign antigens. Mechanisms by which the distinct rearrangements expressed by the hybridomas might have been generated are discussed.

Preferential rearrangement of V kappa 4 gene segments in pre-B cell lines

Examination of the in vitro V kappa gene rearrangements of murine adult bone marrow-derived pre-B cell lines reveals that 21 of 25 (84%) cell lines have rearranged a member of the V kappa 4 family. In contrast, analysis of two V kappa cDNA libraries prepared from LPS-stimulated adult spleen cells indicates that only 17% of the Ig kappa cDNAs contain sequences belonging to the V kappa 4 gene family. Half of the pre-B cell lines examined also share an 8-kbp BamHI reciprocal product (rp). However, these rp do not involve the same V kappa gene, indicating that conserved BamHI sites exist 3' of some V kappa genes. This rp is also readily detected in DNA from normal adult spleen cells, suggesting that the in vitro rearrangements examined in this study are representative of kappa rearrangements that occur in vivo. We suggest that, unlike the diverse V kappa repertoire expressed by mature B cells, the germline V kappa segments involved in initial rearrangements of the Ig kappa locus are highly restricted, and that an initial V kappa 4 rearrangement is probably followed by other, more random recombination events.

Stochastic pairing of heavy-chain and kappa light-chain variable gene families occurs in polyclonally activated B cells

Frequencies of 25 immunoglobulin heavy-chain and kappa light-chain variable (VH + V kappa) gene-family pairings expressed in splenic B-cell populations were determined by hybridization of VH- and V kappa-family-specific DNA probes to mitogen-induced B-cell colonies from C57BL/6 mice or hybridomas derived from BALB/c and NZB mice. Both analyses support the conclusion that VH and V kappa gene families pair without bias; as would be expected for random association, the frequencies of specific VH + V kappa pairs may be estimated by the product of the independent VH and V kappa frequencies. Based upon the frequencies at which 9 VH and 12 V kappa gene families are expressed, we calculated the expected usage for approximately 100 VH + V kappa family pairings in neonatal and adult C57BL/6 mice. Variability in the expression of such VH + V kappa pairings is considerable; pairs representing greater than 10% to less than 0.01% of the splenic B-cell population occur. This variability is most pronounced in the neonate, where 6 VH + V kappa family pairs account for nearly 40% of all mitogen-reactive B cells. As the neonate matures, the distribution of frequencies for VH + V kappa pairings becomes more nearly uniform. This process may underlie the patterned acquisition of humoral immune responsiveness.

Corrective recombination of mouse immunoglobulin kappa alleles in Abelson murine leukemia virus-transformed pre-B cells

Previous characterization of mouse immunoglobulin kappa gene rearrangement products cloned from murine plasmacytomas has indicated that two recombination events can take place on a single kappa allele (R. M. Feddersen and B. G. Van Ness, Proc. Natl. Acad. Sci. USA 82:4792-4797, 1985; M. A. Shapiro and M. Weigert, J. Immunol. 139:3834-3839, 1987). To determine whether multiple recombinations on a single kappa allele can contribute to the formation of productive V-J genes through corrective recombinations, we have examined several Abelson murine leukemia virus-transformed pre-B-cell clones which rearrange the kappa locus during cell culture. Clonal cell lines which had rearranged one kappa allele nonproductively while maintaining the other allele in the germ line configuration were grown, and secondary subclones, which subsequently expressed kappa protein, were isolated and examined for further kappa rearrangement. A full spectrum of rearrangement patterns was observed in this sequential cloning, including productive and nonproductive recombinations of the germ line allele and secondary recombinations of the nonproductive allele. The results show that corrective V-J recombinations, with displacement of the nonproductive kappa gene, occur with a significant frequency (6 of 17 kappa-producing subclones). Both deletion and maintenance of the primary (nonfunctional) V-J join, as a reciprocal product, were observed.

Corrective recombination of mouse immunoglobulin kappa alleles in Abelson murine leukemia virus-transformed pre-B cells

Previous characterization of mouse immunoglobulin kappa gene rearrangement products cloned from murine plasmacytomas has indicated that two recombination events can take place on a single kappa allele (R. M. Feddersen and B. G. Van Ness, Proc. Natl. Acad. Sci. USA 82:4792-4797, 1985; M. A. Shapiro and M. Weigert, J. Immunol. 139:3834-3839, 1987). To determine whether multiple recombinations on a single kappa allele can contribute to the formation of productive V-J genes through corrective recombinations, we have examined several Abelson murine leukemia virus-transformed pre-B-cell clones which rearrange the kappa locus during cell culture. Clonal cell lines which had rearranged one kappa allele nonproductively while maintaining the other allele in the germ line configuration were grown, and secondary subclones, which subsequently expressed kappa protein, were isolated and examined for further kappa rearrangement. A full spectrum of rearrangement patterns was observed in this sequential cloning, including productive and nonproductive recombinations of the germ line allele and secondary recombinations of the nonproductive allele. The results show that corrective V-J recombinations, with displacement of the nonproductive kappa gene, occur with a significant frequency (6 of 17 kappa-producing subclones). Both deletion and maintenance of the primary (nonfunctional) V-J join, as a reciprocal product, were observed.

Autoantibodies and the fetal antibody repertoire

Developing fetal B cells preferentially rearrange a restricted subset of the encoded antibody gene segments. There are striking structural similarities between elements expressed early in man and in mouse, most evident on comparison of murine VH elements from the VH7183 family to human VH elements of the VH3 family. The similarity is pronounced in two framework regions which together encode a possible binding site that is distinct from the classical antigen-combining site. By comparing all known human and murine VH gene sequences, we have demonstrated that these regions have been conserved in a family-specific manner throughout the mammalian radiation. The "non-conserved" spacer of the recombinase recognition signal is also highly conserved in a family-specific manner, suggesting a mechanism by which the expression of family-dependent features may be regulated. The evidence that such features contribute to the high incidence of self- and poly-specificity in the fetal antibody repertoire is reviewed.