Repertoire of human antibodies against the polysaccharide capsule of Streptococcus pneumoniae serotype 6B

We examined the repertoire of antibodies to Streptococcus pneumoniae 6B capsular polysaccharide induced with the conventional polysaccharide vaccine in adults at the molecular level two ways. In the first, we purified from the sera of seven vaccinees antipneumococcal antibodies and determined their amino acid sequences. Their VH regions are mainly the products of VH3 family genes (candidate genes, 3-23, 3-07, 3-66, and 3-74), but the product of a VH1 family gene (candidate gene, 1-03) is occasionally used. All seven individuals have small amounts of polyclonal kappa+ antibodies (Vkappa1 to Vkappa4 families), although kappa+ antibodies are occasionally dominated by antibodies formed with the product of the A27 Vkappa gene. In contrast, lambda+ anti-6B antibodies are dominated by the antibodies derived from one of 3 very similar Vlambda2 family genes (candidate genes, 2c, 2e, and 2a2) and Clambda1 gene product. The Vlambda2(+) antibodies express the 8.12 idiotype, which is expressed on anti-double-stranded-DNA antibodies. In one case, Vlambda is derived from a rarely expressed Vlambda gene, 10a. In the second approach, we studied a human hybridoma (Dob1) producing anti-6B antibody. Its VH region sequence is closely related to those of the 3-15 VH gene (88% nucleotide homology) and JH4 (92% homology). Its VL region is homologous to the 2a2 Vlambda2 gene (91%) and Jlambda1/Clambda1. Taken together, the V region of human anti-6B antibodies is commonly formed by a VH3 and a Vlambda2 family gene product.     

Immunoglobulin repertoire of B lymphocytes infiltrating breast medullary carcinoma

Tumor specific peptides recognized by T lymphocytes infiltrating solid tumors, as well as the corresponding T cell receptor (TcR) repertoire usage, have been extensively investigated. By contrast, tumor infiltrating B cells and their immunoglobulin (Ig) repertoire have been studied only in a limited number of tumors. The objective of the present study was to determine, whether DNA sequence analysis of the expressed immunoglobulin variable regions of B cells that infiltrate breast cancer, could be used to reveal a potential specific tumor binding capacity of the antibodies. To answer this question, about 200 expressed Ig heavy (VH) and light chain variable gene (VL) regions were cloned, sequenced and comparatively analysed from a typical medullary beast carcinoma (MBC), where the massive B and plasma cell infiltration correlates with favourable prognosis despite of its high grade. The tumor infiltrating B cell Ig heavy and light chain sequences could be classified into clusters, families and subgroups, based on the identity level to germline, showing a pattern of oligoclonality. Some overrepresented clusters could be determined. In the course of a detailed analysis and search in Blastn database, a number of VH and VL sequences showed more than 99% homology to DNA sequences of Ig VH region, with proved tumor antigen binding capacity. Our data suggest, that potential tumor binder Ig VH and VL sequences might be selected using a detailed immunoglobulin variable region analysis. This new approach might have a benefit for further antibody engineering, as difficulties in search for tumor binders by phage library selection might be reduced and the time for selection shortened.     

Systematic design and testing of nested (RT-)PCR primers for specific amplification of mouse rearranged/expressed immunoglobulin variable region genes from small number of B cells

The aim of this study was to develop a highly specific and sensitive (RT-)PCR capable of potentially amplifying the rearranged/expressed VH and VL gene belonging to any mouse immunoglobulin V gene family from a single or a small number of B cells. A database of germline immunoglobulin sequences was used to design 112 primers for a nested (RT-)PCR based strategy to cover all VH, VL, JH, JL, CH and CL gene families/genes from C57BL/6 and BALB/c mice. 93.7% of the primers had 4-fold or less, while 71.4% had no degeneracy. The proportions of germline V genes to which the primers bind with no, up to 1 and up to 2 mismatches are 59.7%, 84.1% and 94.9%, respectively. Most but not all V gene family specific primers designed allow amplification of full-length V genes. The nested primers permit PCR amplification of rearranged V genes belonging to all VH and VL gene families from splenocyte genomic DNA. The V gene family-specific nature of the primers was experimentally confirmed for randomly selected 6 VH and 6 Vkappa families, and all Vlambda genes. The broad V gene family coverage of our primer set was experimentally validated by amplifying the rearranged/expressed VH and VL genes from splenocytes and a panel of 38 hybridomas under conditions where primer mixes and genomic DNA or total RNA was used as starting template. We observed no or low-level cross-family priming. Pooled constant region specific primers allowed efficient RT-PCR amplification of H and L chain isotypes. The expressed VH and VL genes belonging to different V gene families RT-PCR amplified from a mixture of hybridomas in a representative manner. We successfully amplified the expressed VH and Vkappa gene from a single hybridoma cell by RT-PCR and from 10-15 microdissected B cells by genomic PCR. This, first of its kind, comprehensive set of highly sensitive and specific nested primers that provide broad V gene family coverage will open up new avenues and opportunities to study various aspects of mouse B cell biology.     

Nucleotide sequence of the VH, VL regions of an anti-idiotopic antibody reacting with a private idiotope of the anti-lysozyme D1.3 antibody

Antibody E225 reacts with a private idiotope of the anti-lysozyme antibody D1.3. A complex between the Fab fragments from these BALB/c monoclonal antibodies has been crystallized and the determination of the three-dimensional structure of this idiotope-anti-idiotope complex is under way. The nucleotide VH and VL sequences of E225 presented here have been determined to provide the amino acid sequence information necessary for the interpretation of the high resolution electron density maps of the complex, obtained by X-ray crystallography. The cDNAs synthesized from the Vkappa and VH mRNAs were cloned in E. coli. Both cDNA strands were sequenced by the dideoxy termination method. The translated amino acid sequence shows that Vkappa, VH correspond to groups five (V) and II(b) of mouse immunoglobulin light and heavy chains, respectively. Sequence alignments between the complementarity determining regions of E225 and the antigenic determinant of lysozyme recognized by D1.3 do not indicate whether or not the anti-idiotopic antibody structurally mimics the external antigen.     

Variable region genes of human monoclonal autoantibodies to histones H2A and H2B from a systemic lupus erythematosus patient

An antibody phage library obtained from peripheral blood lymphocytes of a systemic lupus erythematosus (SLE) patient was used to isolate four monoclonal autoantibodies against histones H2A and H2B. Analysis of the variable region sequences revealed that the anti-histone monoclonal antibodies were not clonally related; they used VH genes from three different VH gene families (VH3, VH4, and VH5) and distant members of the Vkappa group (L25, L6, A27, and O8) in conjunction with different D and J gene segments. These observations suggest that certain gene families or segments are not critical in producing anti-histone autoantibodies in SLE. Most of the utilized VH and Vkappa sequences were highly mutated and the complementarity-determining regions (CDRs) varied greatly in length. The VH region of the antibody SLEhis18 had an isoelectric point of 6.1, and 29% of the mutations were changes to acidic amino acid residues. The second CDR (CDR2) of SLEhis18 VH contained one basic and three acidic residues. Acidic residues were observed in the CDR3 regions of VH, but not VL, in all isolated clones; this is unusual, as most autoantibodies are comprised predominantly of non-acidic residues. This is the first report of a systematic sequence analysis of human anti-histone monoclonal antibodies. Our results suggest that certain V genes are not important for autoreactive specificity to histones in SLE; instead, other mechanisms such as an existence of acidic residues and somatic mutations in CDRs are required for specific binding to histones, which might play a role as a stimulatory autoantigen for the activation of autoantibody-producing B-cells and the selection of high affinity antibody.     

Heterogeneity of cross-reactive idiotypes. Serological and structural analysis of monoclonal anti-p-azobenzene-arsonate antibodies expressing major and minor idiotypes

Hybridoma-derived monoclonal anti-p-azobenzene-arsonate (ABA) antibodies were obtained from fusions of ABA-KLH primed A/J spleen cells with three different myeloma cell lines. Of the 156 antibody secreting hybridomas 24% carried the cross-reactive idiotype (CRI), which is known to be shared by 20-70% of anti-ABA serum antibodies in A/J mice. The isotypes, SDS-PAGE patterns and the partial amino acid sequences of the V-regions of one CRI negative and six CRI positive hybridoma proteins were determined. These antibodies were IgGl, kappa and IgG2b, kappa. Some idiotype carrying monoclonal antibodies appeared to be serologically identical. Although the partial VH amino acid sequences of these monoclonal antibodies showed great homology with each other and with serum antibody, several sequence variations in framework residues as well as in the first and second complementarily determining regions (CDRs) were found. The cross-reactive idiotype of the anti-ABA antibodies, therefore, exhibits structural microheterogeneity, i.e. it consists of a family of non-identical but closely related molecules as previously reported (Alkan et al., 1980; Estess et al., 1980; Marshak-Rothstein et al. 1980). Here the N-terminal sequence of the VH regions from 14 CRI+ and 8 CRI- antibodies as well as the VL regions from 11 CRI+ and 8 CRI- monoclonal antibodies are compared. Analysis of the available data demonstrated that there are pairs of hybridoma proteins (both CRI+ and CRI-) which have identical sequences for VH or VL. This suggests that there exist a minimum of 4 germ line genes coding for CRI+ VH, CRI+ VL, CRI- VH and CRI- VL respectively. In addition, CRI+ VL has always been found in association with a CRI+ VH.     

Molecular studies of anti-HLA-A2 using light-chain shuffling: a structural model for HLA antibody binding

Human leukocyte antigen (HLA) A2 is one of the most immunodominant HLA antigens. Through a process of light-chain variable domain (VL) shuffling, we analyzed the VL domains' role in anti-HLA-A2/A28-binding site diversity. This was achieved by combining a VH3-30-encoded HLA-A2/A28-specific heavy-chain variable domain with 10(4) non-immune VL domains. Twelve HLA-A2/A28-specific antibodies were subsequently identified. VL gene analysis demonstrated an absence of Vlambda domains and that all have VkappaI-encoded light chains. The affinities correlated with the VkappaI gene present, with the seven highest affinity antibodies using Vkappa domains encoded by the O18 gene segment. A 300-fold difference in affinity was observed between the 12 antibodies, and homology modeling demonstrated a correlation between electrostatic surface potential of the antigen-binding site and affinity for HLA. Overlap between the T-cell receptor-binding site and that of the antibodies was indicated by inhibition of cytotoxic T-lymphocyte killing of peptide-pulsed target cells. A model of antibody binding to HLA-A2 suggested contact with both alpha helices of the HLA molecule, such that the antigen-binding site spans the peptide-binding groove. These data increase the understanding of antibody recognition of HLA and may facilitate the production of clonotypic antibodies with peptide-specific binding.     

Analysis of idiotope structure of ovarian cancer antibodies: recognition of the same epitope by two monoclonal antibodies differing mainly in their heavy chain variable sequences.

Two MoAbs, independently raised against ovarian carcinoma cells and referred to as OV-TL3 and OV-TL16, display an identical reaction pattern with a membrane-associated protein in both normal and malignant ovarian cells. Also, a similar binding affinity constant and a similar number of binding sites per cell indicate that both MoAbs bind to the same antigen. Competition assays reveal that OV-TL16 is able to compete with OV-TL3 for binding to OVCAR-3 cells. Epitope mapping using a filamentous phage hexapeptide epitope library showed that both MoAbs are able to select identical phages, suggesting that their epitopes are identical or at least overlapping. However, purified polyclonal and monoclonal anti-idiotypic antibodies directed against OV-TL3 failed to recognize the OV-TL16 idiotype, indicating that the structure of the antigen-binding regions of both antibodies is distinct. This was corroborated by molecular cloning and sequencing of the variable heavy (VH) and light (VL) chain immunoglobulin regions of both MoAbs. The VH regions of both antibodies were found to be distinct, whereas the VL regions are almost identical. Computer modelling of the idiotypes suggests that the complementarity determining regions (CDR), with the exception of VHCDR3, have (almost) identical spatial configurations. Our data indicate that, although structurally different in their VH regions, OV-TL3 and OV-TL16 are able to bind to identical epitopic regions on the antigen, because differences in primary structure do not exclude the formation of sufficient and similar spatial structures for the interaction with an epitope.     

The antibody repertoire in evolution: chance, selection, and continuity

All jawed vertebrates contain the genetic elements essential for the function of the adaptive/combinatorial immune response, have diverse sets of natural antibodies resulting from segmental gene recombination, express comparable functional repertoires and can produce specific antibodies following appropriate immunization. Profound variability occurs in the third hypervariable (CDR3) segments of light and heavy chains even within antibodies of the same ostensible specificity. Germline VH and VL elements, as well as the joining (J) segments are highly conserved among the distinct vertebrate species. Conservation is particularly noted among the VH3-like sequences of all jawed vertebrates in the FR2 and FR3 segments, as well as in the FGXGT(R or K)L J-segment characteristic of light chains and TCRs and the WGXGT(uncharged)VT JH segments. Human VH3-53 and Vlambda6 family orthologs may be present over the entire range of vertebrates. Models of the three-dimensional structures of shark VH/VL combining sites indicate similarity in framework structure and comparable CDR usage to those of man. Although carcharhine shark VH regions show greater than 50% identity to the human VH germline prototype, searches of lower deuterostome and invertebrate databases fail to detect molecules with significant relatedness. Overall, antibodies of jawed vertebrates show tremendous individual diversity, but are constructed incorporating design features that arose with the evolutionary emergence of the jawed vertebrates and have been conserved through at least 450 million years of evolutionary time.     

The role of CD40-CD40 ligand (CD154) interactions in immunoglobulin light chain repertoire generation and somatic mutation.

To determine whether CD40 ligation influences the molecular and selective mechanisms that govern the development of the human Ig light chain repertoire, analysis of the Vkappa and Vlambda repertoires of CD19+ B cells obtained from a patient with X-linked hyper IgM syndrome (XHIM) and a nonfunctional CD154 was carried out. The nonproductive Vkappa and Vlambda repertoires were largely comparable to that of the normals with respect to V gene and J segment distribution as well as CDR3 length and VLJL joint complexity. Comparison of the nonproductive and productive repertoires indicated that a limited number of VL genes were positively and negatively selected in the XHIM patient. Although mutations were observed in the XHIM VL repertoires, the frequency of mutations was significantly lower than in normals. Typical targeting of these mutations into RGYW/WRCY motifs was significantly reduced and subsequent selection of RGYW/WRCY mutations, which is normally observed, was not found. These results indicate that CD40 ligation is not required for generation of the light chain repertoire, positive selection of some Vk rearrangements, negative selection of specific VL genes, and some degree of somatic mutation. Importantly, however, targeting of mutations to RGYW/WRCY motifs and subsequent selection of these mutated motifs does not occur in the absence of CD40 ligation.     

Construction and molecular characterization of mouse single-chain variable fragment antibodies against Burkholderia mallei and Burkholderia pseudomallei

We have selected two lipopolysaccharide (LPS) specific Burkholderia mallei mouse monoclonal antibodies (mAbs) and four anti-capsular B. pseudomallei-specific mAbs to generate mouse single-chain variable fragment (scFv) antibodies. This selection was made through extensive in vitro and in vivo assay from our library of mAbs against B. mallei and B. pseudomallei. We initially generated the mouse immunoglobulin variable heavy chain (VH) and light chain (VL) regions from each of these six selected mAbs using a phage display scFv technology. We determined the coding sequences of the VH and VL regions and successfully constructed two B. mallei-specific scFv phage antibodies consisting of two different VH (VH1 and VH2) and one Vλ1 families. Four scFvs constructed against B. pseudomallei had two VH (VH1 and VH6) and two VL (Vκ4/5 and Vκ21) families. All of six scFv antibodies constructed demonstrated good binding activity without any rounds of biopanning against B. mallei (M5D and M18F were 0.425 and 0.480 at OD405nm) and B. pseudomallei (P1E7, P2I67, P7C6, and P7F4 were 0.523, 0.859, 0.775, and 0.449 at OD405nm) by ELISA, respectively. A comparison of the immunoglobulin gene segments revealed that the gene sequences in complementarity-determining regions (CDRs) of three out of four B. pseudomallei-specific scFvs are highly conserved. We determined that the two B. mallei-specific scFvs have different CDRs in the VH, but the amino acid sequences of CDRs in the VL are conserved. This high sequence homology found in CDRs of VH or VL of these mAbs contributes to our better understanding and determination of binding to the specific antigenic epitope(s). The scFv phage display technology may be a valuable tool to develop and engineer mAbs with improved antigen-binding affinity.     

Chicken IgL variable region gene conversions display pseudogene donor preference and 5' to 3' polarity

Chicken immunoglobulin variable region diversity is generated during B-cell development in the bursa of Fabricius by intrachromosomal gene conversion, resulting in the replacement of sequence blocks within the unique rearranged VL1 and VH1 genes with homologous sequences derived from V region pseudogene segments (psi V). In this report, the nucleotide sequences of 217 gene conversion events in 52 random IgL clones were analyzed to characterize the molecular mechanism of gene conversion. The frequency of psi VL usage as gene conversion donors is shown to depend on the proximity of the psi VL segment to VL1, extent of homology with VL1, and relative orientation of the psi VL segments. Gene conversion events are not observed in the 5' region of homology between psi VL segments and VL1, but are distributed throughout the remainder of the VL1 exon. The 5' ends of individual gene conversion events always begin in regions of homology between the donor psi VL and recipient VL1 gene, whereas the 3' ends can occur in regions of nonhomology and often have nucleotide insertions or deletions. These results suggest a 5' to 3' polarity in the gene conversion mechanism. The implications of our data are discussed in relation to current molecular models of gene conversion.     

Molecular analysis of cross-reactive anti-myosin/anti-streptococcal mouse monoclonal antibodies

Nucleotide sequences of VH- and VL-genes of anti-myosin/anti-streptococcal monoclonal antibodies (mAbs) were analyzed and compared with their highly detailed antigen binding reactivities. Antigen-specificities of the cross-reactive mAbs included myosin, streptococcal M-protein, actin, keratin, N-acetyl-beta-D-glucosamine, vimentin, DNA, tropomyosin, troponin, and laminin as previously described. After nucleotide sequence analysis, homology indicated that some of the V gene sequences aligned with antibodies recognizing gangliosides and blood group antigens glycophorin M and N. Therefore, mAb reactivity with gangliosides and glycophorin M and N was identified. The cross-reactive mAbs utilized a heterogeneous group of germline V-heavy genes comprised of nine J558-, four 7183- and two Q52-family VH-genes. Germline V-light genes utilized by the mAbs included six Vkappa4/5-, three Vkappa8-, two Vkappa10-, three Vkappa19- and one Vkappa23-family VL-genes. No preferential VH/VL-chains correlated with any of the 12 different antigen reactivities, even for mAbs with nearly identical cross-reactivities. However, we did find that the cross-reactive mAb germline genes within a V gene family shared more homology among themselves than with other germline genes within their V gene families, suggesting convergent mutation. Cross-reactive mAbs with the highest relative avidity for myosin were found in the VH7183 family which contained two cytotoxic mAbs. Antibodies with V gene sequences most homologous to those of our cross-reactive anti-myosin/anti-streptococcal mAbs had specificities for laminin, DNA, carbohydrates, or blood group antigens and were reported to cause autoimmune disease in mice.     

Three monoclonal immunoglobulins, an IgG2(kappa), an IgM(kappa) and an IgM/A hybrid, in one patient. II. Sharing of common variable regions

Amino acid sequencing and haemagglutination inhibition studies were performed on three monoclonal immunoglobulins (an IgG2, an IgM and an IgM/A hybrid) isolated from a patient afflicted with a multiple gammopathy. The results demonstrated that all three proteins have shared idiotypic determinant(s). Furthermore, the light chains of all three paraproteins have identical NH2-terminal amino-acid sequences at all positions determined thus far. Similarly, the NH2-terminal amino-acid sequence of the gamma 2 chains is identical to that of the mu chain. The evidence strongly suggests a common ancestral clonal origin for cells which produce these paraproteins and that identical variable region (VH and VL) genes were used by the ESM lymphocyte subclones in the biosynthesis of the respective IgM, IgG, and IgM/A hybrid molecules. The occurrence of a mu/alpha hybrid chain, which shares identical V regions with a mu and a gamma chain, is consistent with the concept that IgM-producing cells can develop directly into cells producing other classes of immunoglobulins via separate pathways during B-cell maturation.     

Rheumatoid factor autoantibody-binding site: a molecular analysis using monoclonal antibodies with dual anti-TNP and anti-IgG activities

Two out of five murine IgG3 anti-trinitrophenyl (TNP) monoclonal antibodies (mAb) obtained either by immunization with TNP-keyhole limpet hemocyanin (KLH) (CB1, CB5, CB6 and 4H10) or with dinitrophenyl-lipopolysaccharide (9A6), exhibited anti-IgG rheumatoid factor (RF) activity (CB6 and 4H10). The anti-IgG activity of these two anti-TNP RF was specifically inhibited by murine IgG as well as by the hapten TNP. In order to identify the structural basis for the anti-IgG activity, the nucleotide sequences encoding the VH and VL regions were determined. By comparing the V regions of the non-RF and RF anti-TNP mAb, it was found that one anti-TNP RF antibody, CB6, shares virtually identical VL and VH regions with two anti-TNP antibodies, CB1 and CB5, but markedly differs from these in the D region. Furthermore, the light chain framework region 2 (FR2) and FR3 of non-RF mAb, CB1, CB5 and 9A6, have amino acid sequences almost identical to those claimed for anti-IgG1 RF activity (Shlomchik et al., J. Exp. Med. 1986. 164: 407). Our findings suggest, at least in the case of CB6 mAb, the involvement of CDR, but not light chain FR residues, in IgG binding.     

Pertinence of kappa and lambda recombinant antibodies directed against thyroid peroxidase in thyroid autoimmune disease

Forty-one single-chain variable region fragments (scFvs) directed against thyroid peroxidase (TPO) were obtained by phage display libraries constructed from thyroid-infiltrating B cells of Graves' disease patients. Among these scFvs, 24.4% used a Vkappa light chain whereas 75.6% shows a light chain of Vlamda origin. Study of light chain gene usage in the TPO antibody repertoire demonstrated a dominance of the Vkappa 1-39 and Vlambda 1-51 genes. Thyroid peroxidase probing of overlapping peptides covering the amino acid sequences of anti-TPO T2/kappa and T13/lambda variable regions demonstrated a more restricted antigen recognition on T13/lambda than on T2/kappa. These two recombinant antibodies, expressed as whole IgG1 in the baculovirus/insect cell system, inhibited the binding to TPO of serum TPO autoantibodies whatever the light chain. Our study indicates that lambda as well as kappa light chain usage are found in the TPO antibody repertoire of thyroid-infiltrating B cells and are pertinent in the pathogenesis of autoimmune thyroid disease.     

Fundamental characteristics of the expressed immunoglobulin VH and VL repertoire in different canine breeds in comparison with those of humans and mice

Complementarity determining regions (CDR) are responsible for binding antigen and provide substantial diversity to the antibody repertoire, with VH CDR3 of the immunoglobulin variable heavy (VH) domain playing a dominant role. In this study, we examined 1200 unique canine VH and 500 unique variable light (VL) sequences of large and small canine breeds derived from peripheral B cells. Unlike the human and murine repertoire, the canine repertoire is heavily dominated by the Canis lupus familiaris IGHV1 subgroup, evolutionarily closest to the human IGHV3 subgroup. Our studies clearly show that the productive canine repertoire of all analyzed breeds shows similarities to both human and mouse; however, there are distinct differences in terms of VH CDR3 length and amino acid paratope composition. In comparison with the human and murine antibody repertoire, canine VH CDR3 regions are shorter in length than the human counterparts, but longer than the murine VH CDR3. Similar to corresponding human and mouse VH CDR3, the amino acids at the base of the VH CDR3 loop are strictly conserved. For identical CDR positions, there were significant changes in chemical paratope composition. Similar to human and mouse repertoires, the neutral amino acids tyrosine, glycine and serine dominate the canine VH CDR3 interval (comprising 35%) although the interval is nonetheless relatively depleted of tyrosine when compared to human and mouse. Furthermore, canine VH CDR3 displays an overrepresentation of the neutral amino acid threonine and the negatively charged aspartic acid while proline content is similar to that in the human repertoire. In general, the canine repertoire shows a bias towards small, negatively charged amino acids. Overall, this analysis suggests that functional canine therapeutic antibodies can be obtained from human and mouse sequences by methods of speciation and affinity maturation.