Structural characterization of idiotopes by using antibody variants generated by site-directed mutagenesis

Four anti-idiotopic mAB, 107, MB, AI, and AD8, react with mouse hybridoma protein 36-65 specific for the hapten p-azophenylarsonate. The four antiidiotypic antibodies do not react with hybridoma protein 36-71, a somatically mutated variant of 36-65 whose H and L chain V region sequence differs at 19 amino acid positions. To determine which regions of 36-65 are important for the interaction with each of the four anti-idiotypic antibodies, variants of 36-65 containing one or more of the 36-71 substitutions were generated by oligonucleotide-directed mutagenesis of the rearranged 36-65 H chain V region gene, followed by expression of mutant proteins containing either the 36-65 or the 36-71 L chain in transfected hybridoma cells. Idiotypic characterization of the mutant proteins showed that reactivity correlates with the 36-65 H chain, but some contributions from the 36-65 L chain come into play. In the 36-65 H chain V region, idiotopes were mapped to the first and third complementarity-determining regions for anti-idiotypic antibodies 107, MB, and AI, and to all three complementarity-determining regions for anti-idiotypic antibody AD8. The binding of all four anti-idiotypic antibodies to hybridoma protein 36-65 was hapten inhibitable. However, a comparison between the effect of individual 36-71 substitutions on idiotope expression and their effect on Ag-binding affinity suggests that none of the four anti-idiotypic antibodies bodies mimics the structure of Ag.     

Immunoglobulin idiotype and anti-anti-idiotype utilize the same variable region genes irrespective of antigen specificity

Idiotypic determinants characterizing certain antibody specificities have been proven valuable structural and genetic markers in studies of antibody diversity and regulation. The heritable predominant idiotype associated with the response to p-azophenylarsonate in A/J mice consists of a set of highly homologous (greater than 95%) heavy and light chain variable region amino acid sequences probably arising by somatic mutation from one or a few V region genes. We examined a peculiar set of monoclonal antibodies that have been defined as CRI by serologic analysis, but that have no affinity for the hapten Ars. These antibodies were elicited by immunization with anti-CRI rather than by the conventional immunization with antigen. The amino acid sequences of the amino terminal half of the V regions of these anti-(anti-CRI) antibodies are indistinguishable from those of conventional Ars-binding CRI antibodies. Thus, Ars-binding CRI and Ars-nonbinding anti-(anti-CRI) are derived from similar or identical VH and VL genes.     

Three-dimensional structure determination of an anti-2-phenyloxazolone antibody: the role of somatic mutation and heavy/light chain pairing in the maturation of an immune response.

The three-dimensional structure of the Fab fragment of an anti-2-phenyloxazolone monoclonal antibody (NQ10/12.5) in its native and complexed forms has been determined at 2.8 and 3.0 A resolution, respectively. Identification of hapten-contacting residues has allowed us to evaluate the contribution of individual somatic point mutations to maturation of the immune response. In particular, amino acid residues 34 and 36 of the light chain, which are frequently mutated in antibodies with increased affinity for 2-phenyloxazolone, are shown to interact directly with the hapten. We propose that the strict maintenance of certain amino acid sequences at the potentially highly variable VL-JL and VH-D-JH junctions observed among anti-2-phenyloxazolone antibodies is due largely to structural constraints related to antigen recognition. Finally, the three-dimensional model of NQ10/12.5, which uses the typical light chain of primary response anti-2-phenyloxazolone antibodies but a different heavy chain, allows an understanding of how, by preserving key contact residues, a given heavy chain may be replaced by another, apparently unrelated one, without loss of hapten binding activity and why the V kappa Ox1 germline gene is so frequently selected amongst the other known members of this family.     

Mutational analysis of the affinity maturation of antibody 48G7

The affinity maturation of antibody 48G7 from its germline predecessor 48G7g has been studied at a molecular level through a combination of structural and biochemical means. Each of the nine somatic mutations accumulated during affinity maturation has been assessed for gain or loss of function in both the germline and affinity-matured antibodies. Individual somatic mutations were found to be either positive or neutral in their effects on affinity for hapten JWJ1, with a marked context-dependence for some sites of mutation. In a number of cases significant cooperativity was found between pairs of somatically mutated residues. Interpretation of the structural changes introduced by many of the point mutations has been possible due to the availability of high-resolution crystal structures of 48G7g and 48G7, and mechanisms by which these structural changes may result in enhanced affinity for hapten have been identified. Precise dissection of structure-function relationships in this system provides additional insights into the role of cooperativity in the evolution of antibody affinity. Comparison of 48G7 with previously characterized systems provides a varied view of the structure-function mechanisms by which the humoral immune system produces large increases in affinity.     

Sequence and fine specificity analysis of primary 511 anti-phosphorylcholine antibodies

The primary antibody response of mice to phosphorylcholine (PC) is dominated by antibodies using the T15 L chain. Anti-PC antibodies using the 511 L chain are prominent only in secondary responses to PC coupled to proteins, are somatically mutated, and all have an extra amino acid at the Vh-D junction, compared with T15 antibodies. The aim of the experiments reported here was to determine if the extra junctional amino acid alone was sufficient to generate a 511 PC-binding antibody, or if somatic mutation or other junctional changes were also necessary. We also wished to determine if unmutated 511 antibodies had sufficient affinity for PC to appear in the primary response. To increase the frequency of primary 511 antibodies, we generated a series of hybridomas from M167 L chain transgenic mice immunized 4 days earlier with either Streptococcus pneumonia R36a or PC-keyhole limpet hemocyanin (KLH). We determined the relative affinity of the antibodies, and sequenced their H chain V regions. The results showed that: 1) somatic mutations are not required for 511 antibodies to bind PC; 2) primary 511 antibodies all had lower relative affinities for PC than T15 while having similar affinities to T15 for TNP-aminophenyl PC, and higher affinities for the PC analogs nitrophenyl PC and choline; 3) all antibodies had the 511-specific insertion of an extra amino acid, usually Ala, at the VhD junction, compared with T15; 4) immunization with R36a, but not PC-hemocyanin, elicited antibodies with a specific Tyr----Asp substitution in the D region, indicating Ag selection based on fine specificity differences; 5) the total length of CDR3 was conserved in most anti-PC-hemocyanin antibodies, whereas the anti-R36a antibodies predominantly had longer CDR3 sequences; and 6) there were unique substitutions in most antibodies, including significant sequence heterogeneity in the D-Jh junction. We conclude that Ag selection on the basis of affinity for PC biases the primary anti-PC response in favor of T15, and that 511 precursors with their alternative fine specificities contribute the precursors that are expanded in the secondary anti-PC-KLH responses.     

Mimicking Somatic Hypermutation: Affinity Maturation of Antibodies Displayed on Bacteriophage Using a Bacterial Mutator Strain

Human antibodies can now be isolated from antibody repertoires displayed on the surface of filamentous bacteriophage in a process that mimics the primary immune response. Here we have attempted to mimic the secondary response, the natural process of affinity maturation of antibodies occurring in germinal centres, by multiple cycles of random mutation and selection. Phage displaying a human antibody fragment recognising the hapten 2-phenyl-5-oxazolone were grown in a mutator strain of bacteria (Escherichia coli: mutD5) to generate a large repertoire of antibodies that should include the majority of possible single nucleotide point mutations. The repertoire of phage antibody mutants was then selected by binding to hapten. By multiple rounds of growth in the mutator strain, and increasingly stringent selection, we succeeded in isolating mutants with improved binding affinities; furthermore, the distribution of mutations and nucleotide substitution preferences strongly resembled those of somatic hypermutation. We then constructed a genealogical tree from the sequences of mutants taken at different rounds, and identified four sequentially acquired mutations that together improve the binding affinity of the antibody by a factor of 100-fold (fromK_d320 nM to 3.2 nM).     

Structural basis of affinity maturation of the TEPC15/Vkappa45.1 anti-2-phenyl-5-oxazolone antibodies

Affinity maturation is a process that leads to the emergence of more efficient antibodies following initial antigen encounter and represents a key strategy of the adaptive immunity of vertebrate organisms. Earlier and detailed sequence studies of the antibody response to a model antigen, the hapten 2-phenyl-5-oxazolone (phOx), define three different classes of antibodies. Class I antibodies use the V(H)Ox1/V(kappa)Ox1 gene pair and dominate the early stages of the anti-phOx response, class II antibodies use the V(kappa)Ox1 gene but a different V(H) segment and are common in the intermediate stages, and class III antibodies use the TEPC15/V(kappa)45.1 genes and play the greatest role in the late stages. Only the crystal structure of one anti-phOx antibody, the class II NQ10/12.5 Fab fragment, has been described. Here we report the crystal structures of the scFv form of the low and high affinity anti-phOx class III antibodies NQ10/1.12 and NQ16/113.8 complexed with the hapten. The two antibodies differ by nine amino acid substitutions, all located in the V(H) domain. Analysis of the two structures shows that affinity maturation results from an increase in surface complementarity, as a consequence of a finely tuned and highly concerted process chaperoned by the somatic mutations, and implies a more efficient hapten-induced fit in the mature antibody. The data also demonstrate that class III antibodies respond in a completely different way to the architectural problem of binding phOx compared to the class II antibody NQ10/12.5.     

Clustered H chain somatic mutations shared by anti-p-azophenylarsonate antibodies confer enhanced affinity and ablate the cross-reactive idiotype

A cluster of four consecutive CDR2 somatic mutations are shared by the VH regions of two independently isolated hybridoma antibodies with specificity for p-azophenylarsonate (Ars). The mutations appear to be derived by a series of independent events. To assess the influence of these shared somatic mutations on antibody affinity for Ars and on idiotypy, we introduced them, via site-directed mutagenesis, into the V region of an anti-Ars antibody that was otherwise unmutated and we eliminated them from the mutated context of one of the two antibodies in which they were originally found. Results of affinity measurements by fluorescence quenching and of idiotypic binding assays performed on these engineered mutants demonstrated that the shared mutations increased affinity for Ars and eliminated the predominant Id associated with strain A anti-Ars antibodies and four of five idiotypes defined by anti-idiotypic mAb. These results support the interpretation that a strong affinity-based selection pressure has favored the clonal expansion of B cells with receptors containing these mutations despite the loss of a predominant Id. Thus, in producing antibodies containing V regions conferring high affinity for Ag, the combined processes of somatic mutation and clonal selection have generated a common structural solution through parallel repeats.     

Clonal recruitment and somatic mutation in the generation of immunological memory to the hapten NP.

The nucleotide sequences of the variable regions of lambda 1 chain bearing anti-NP antibodies from the secondary response of C57BL/6 mice were determined. The data indicate that the V186.2 VH gene which dominates the primary anti-NP response is expressed in nine out of 10 secondary response antibodies and is extensively mutated. In the V lambda 1 regions somatic mutations are less frequent. While point mutations predominate, there is suggestive evidence for two conversion events, one involving a one-codon deletion. Most, but not all, secondary response antibodies have a higher affinity (up to 10-fold) for the hapten than is seen in the primary response. The increase in affinity correlates with 'parallel' mutations in CDRs of H and L chains, likely to play a role in hapten binding. The analysis of VDJH rearrangements demonstrates that the secondary response lambda 1 chain-bearing antibodies are produced by a diverse set of B cell clones, which are only rarely expressed in primary responses. These clones are characterized by N-sequence-mediated heterogeneity in the 3' half of CDR3, where the germ line sequence of the D element DFl16.1 predominates in primary response antibodies. The antibodies analyzed in this and in previous work were isolated from idiotypically suppressed mice in order to evaluate whether, intraclonally, idiotype suppression selects antibody mutants into the memory pool, through suppression of the wild-type. A selection of this type was not detectable. However, idiotype suppression may control the pattern of clonotypes expressed in the primary versus the secondary response.     

一种构建改形单域抗体的方法

为验证一种构建改形单域抗体的实用新方法,与以往方法不同的是,该方法不需要对抗体进行空间结构模拟,以确定人源抗体的FRs接受序列及在人源FRs接受序列中哪些氨基酸残基需要突变,并且该方法将抗体的改形与亲和力成熟于同一过程完成,利用该方法构建了改形抗CD28重链单域抗体,根据一种鼠源抗CD28重链单域抗体的氨基酸序列,于GenBank中查得两条与之最同源的人源抗体序列,利用其中一条的FRs作为改形抗体的主框架进行改形构建,将鼠源抗体的CDR区插入到人源FR区后,对人源FR区的一些氨基酸残基进行替换突变,替换的氨基酸残基数及替换原则主要是根据对查到的人源抗体序列,鼠源抗体序列,以及这些序列与Kabat分类中的种属序列进行的比较,为了增加改形抗体基因的多样性,对要被替换的氨基酸残基在基因合成中采用简并的方式,使要被替换的氨基酸残基和替换的氨基酸残基都有机会出现,二者出现的几率各为50%,同时,在将大小不同的合成核苷酸片段采用重叠PCR扩增以获得完整改形抗体基因时,采用高Mg^2 浓度下和使用TaqDNA聚合酶,以进一步随机引入突变,利用重叠PCR产物构建了一个噬菌体抗体库,经过3轮淘选后,获得了几个具有较高免疫学活性的改形抗体,对其中的两个抗体进行了进一步研究,将两个抗体的基因在大肠杆菌BL21（DE3）中表达,复性后的表达蛋白仍具有较高的免疫学活性,结果表明该方法是有效可行的。     

Pushing antibody-based labeling systems to higher sensitivity by linker-assisted affinity enhancement

The sensitivity of antibody/hapten-based labeling systems is limited by the natural affinity ceiling of immunoglobulins. Breaking this limit by antibody engineering is difficult. We thus attempted a different approach and investigated if the so-called bridge effect, a corecognition of the linker present between hapten and carrier protein during antibody generation, can be utilized to improve the affinity of such labeling systems. The well-known haptens 2,4-dinitrophenol (2,4-DNP) and 2,4-dichlorophenoxyacetic acid (2,4-D) were equipped with various linkers, and the resulting affinity change of their cognate antibodies was analyzed by ELISA. Anti-2,4-DNP antibodies exhibited the best affinity to their hapten when it was combined with aminobutanoic acid or aminohexanoic acid. The affinity of anti-2,4-D antibodies could be enhanced even further with longer aliphatic spacers connected to the hapten. The affinity toward aminoundecanoic acid-2,4-D derivatives, for instance, was improved about 100-fold compared to 2,4-D alone and yielded detection limits as low as 100 amoles of analyte. As the effect occurred for all antibodies and haptens tested, it may be sensible to implement the bridge effect in future antibody/hapten-labeling systems in order to achieve the highest sensitivity possible.     

Structural plasticity and the evolution of antibody affinity and specificity

The germline precursor to the ferrochelatase antibody 7G12 was found to bind the polyether jeffamine in addition to its cognate hapten N-methylmesoporphyrin. A comparison of the X-ray crystal structures of the ligand-free germline Fab and its complex with either hapten or jeffamine reveals that the germline antibody undergoes significant conformational changes upon the binding of these two structurally distinct ligands, which lead to increased antibody-ligand complementarity. The five somatic mutations introduced during affinity maturation lead to enhanced binding affinity for hapten and a loss in affinity for jeffamine. Moreover, a comparison of the crystal structures of the germline and affinity-matured antibodies reveals that somatic mutations not only fix the optimal binding site conformation for the hapten, but also introduce interactions that interfere with the binding of non-hapten molecules. The structural plasticity of this germline antibody and the structural effects of the somatic mutations that result in enhanced affinity and specificity for hapten likely represent general mechanisms used by the immune response, and perhaps primitive proteins, to evolve high affinity, selective receptors for so many distinct chemical structures.     

Experimental analysis by site-directed mutagenesis of somatic mutation effects on affinity and fine specificity in antibodies specific for lysozyme.

To experimentally examine the functional roles of somatically derived structural variation in the lysozyme-binding mAb HyHEL-10, we have introduced three different point mutations and one insertion at two different sites in HyHEL-10 by site-directed mutagenesis and expression of the mutant antibodies. Mutation of Asp----Ala at position 101 of the H chain returns a somatically mutated residue to its germline sequence for HyHEL-10, and reduces affinity for chicken lysozyme by approximately 9000-fold. Lengthening the third H chain hypervariable region by two amino acids reduces affinity by about 2000-fold. Two mutations, Asp----Thr at position 101 in the H chain and Lys----Thr at position 49 in the L chain, model somatic differences found in another structurally related but functionally distinguishable mAb and minimally decrease affinity for chicken lysozyme. The H chain mutation Asp101VVH----Thr has little effect on affinity for other avian lysozymes but does alter relative fine specificity for these lysozymes. The L chain mutation Lys49VK----Thr increases affinity for duck lysozyme by approximately fivefold. Neither of the positions mutated, 101 in the H chain nor 49 in the L chain, nor the residues near the insertion contact lysozyme in the x-ray structure of the HyHEL-10 F(ab)-HEL complex. The results suggest that these mutations, which model observed somatic mutations, produce functional variation by indirect or long-range effects.     

Mutational analysis of the cross-reactive idiotype of the A strain mouse.

The elucidation of the structural basis for expression of the cross-reactive Id of the A strain mouse (CRIA) in response to the hapten p-azophenylarsonate has been the object of considerable research effort. Most conclusions regarding the amino acids involved in Id expression have been inferential, based on comparisons of amino acid sequences, chain recombination experiments, or chemical modification of particular amino acids. To more rigorously designate the amino acids critical to the phenotype of the CRIA, a system for the expression and directed mutation of antibody molecules was developed. Based on the baculovirus expression of foreign proteins in cultured insect cells, functional antibodies can be produced at very high levels in vitro. By the process of oligonucleotide-directed mutagenesis, a series of specific amino acid changes were introduced into both the H and L chains of a prototype CRIA expressing antibody molecule. Analysis of the gain or loss of polyclonal Id and each of several monoclonal idiotopes has allowed us to identify more precisely the amino acids responsible for expression of the CRIA. Thus we have shown that expression of the CRIA is equally dependent on amino acids in the second and third CDR of the H chain. Furthermore, the idiotopes studied surround the Ag-binding site and clearly involve multiple interactions in several of the L and H chain CDR.     

N-terminal mutations in the anti-estradiol Fab 57-2 modify its hapten binding properties

Recombinant antibodies often contain N-terminal mutations arising from the use of degenerate cloning primer sets and/or the introduction of restriction sites in the framework 1 regions. We studied the effects of such mutations in a recombinant anti-estradiol Fab fragment derived from the hybridoma cell line 57-2. The 5' ends of the heavy and light chain genes were originally modified to introduce the restriction sites XhoI and SacI, respectively, for cloning purposes. However, the affinity and specificity of the recombinant Fab were lowered compared to the proteolytic Fab' fragment of the parental hybridoma IgG. Replacing the mutated sites with authentic amino acid coding sequences restored the binding properties as well as increased the bacterial production levels fivefold and 10-fold at 30 and 37 degrees C, respectively. Local changes in the antigen binding site were probed by determining the affinity constants (Kd) for estradiol and four related steroids. It was found that the mutated heavy chain amino terminus specifically increased the Kd for testosterone whereas the mutated light chain amino terminus decreased the Kd for all of the steroids to the same extent; the heavy and light chain effects were additive. Analysis of a newly determined crystal structure of the authentic Fab 57-2 in complex with estradiol suggests that mutations in the residue 2 in V(H), and 2 and 4 in the V(L) domain were those responsible for the observed effects. Their general roles as structure-determining residues for the CDR3 loops imply that similar effects can occur with other recombinant antibodies as well.     

Human Germline Antibody Gene Segments Encode Polyspecific Antibodies

Structural flexibility in germline gene-encoded antibodies allows promiscuous binding to diverse antigens. The binding affinity and specificity for a particular epitope typically increase as antibody genes acquire somatic mutations in antigen-stimulated B cells. In this work, we investigated whether germline gene-encoded antibodies are optimal for polyspecificity by determining the basis for recognition of diverse antigens by antibodies encoded by three V_H gene segments. Panels of somatically mutated antibodies encoded by a common V_H gene, but each binding to a different antigen, were computationally redesigned to predict antibodies that could engage multiple antigens at once. The Rosetta multi-state design process predicted antibody sequences for the entire heavy chain variable region, including framework, CDR1, and CDR2 mutations. The predicted sequences matched the germline gene sequences to a remarkable degree, revealing by computational design the residues that are predicted to enable polyspecificity, i.e., binding of many unrelated antigens with a common sequence. The process thereby reverses antibody maturation in silico. In contrast, when designing antibodies to bind a single antigen, a sequence similar to that of the mature antibody sequence was returned, mimicking natural antibody maturation in silico. We demonstrated that the Rosetta computational design algorithm captures important aspects of antibody/antigen recognition. While the hypervariable region CDR3 often mediates much of the specificity of mature antibodies, we identified key positions in the V_H gene encoding CDR1, CDR2, and the immunoglobulin framework that are critical contributors for polyspecificity in germline antibodies. Computational design of antibodies capable of binding multiple antigens may allow the rational design of antibodies that retain polyspecificity for diverse epitope binding.     

Tailored amino acid diversity for the evolution of antibody affinity

Antibodies are a unique class of proteins with the ability to adapt their binding sites for high affinity and high specificity to a multitude of antigens. Many analyses have been performed on antibody sequences and structures to elucidate which amino acids have a predominant role in antibody interactions with antigens. These studies have generally not distinguished between amino acids selected for broad antigen specificity in the primary immune response and those selected for high affinity in the secondary immune response. By studying a large data set of affinity matured antibodies derived from in vitro directed evolution experiments, we were able to specifically highlight a subset of amino acids associated with affinity improvements. In a comparison of affinity maturations using either tailored or full amino acid diversification, the tailored approach was found to be at least as effective at improving affinity while requiring fewer mutagenesis libraries than the traditional method. The resulting sequence data also highlight the potential for further reducing amino acid diversity for high affinity binding interactions.     

Mutational analysis of arsonate binding by a CRIA+ antibody. VH and VL junctional diversity are essential for binding activity

To assess the impact of various heavy and light chain mutations on p-azophenylarsonate binding, murine antibodies have been produced in insect cells (SF9) utilizing a baculovirus expression system. When expressed in this system, an antibody composed of a canonical CRIA+ heavy and light chain can bind antigen and express idiotype indistinguishably from analogous hybridoma-derived antibodies. Antibodies comprised of either light chains mutant at the V-J junction or heavy chains mutant at the V-D junction were found to be incapable of binding arsonate. In addition, substitutions in the first and second complementarity determining regions of the heavy chain were shown to play a role in arsonate binding, most likely related to affinity maturation targeted at the carrier protein. These results confirm the obligatory role that junctional diversity plays in the generation of arsonate-specific antibodies, as well as extend our understanding of the role of other variable region amino acids in arsonate binding.     

Tailored amino acid diversity for the evolution of antibody affinity

Antibodies are a unique class of proteins with the ability to adapt their binding sites for high affinity and high specificity to a multitude of antigens. Many analyses have been performed on antibody sequences and structures to elucidate which amino acids have a predominant role in antibody interactions with antigens. These studies have generally not distinguished between amino acids selected for broad antigen specificity in the primary immune response and those selected for high affinity in the secondary immune response. By studying a large data set of affinity matured antibodies derived from in vitro directed evolution experiments, we were able to specifically highlight a subset of amino acids associated with affinity improvements. In a comparison of affinity maturations using either tailored or full amino acid diversification, the tailored approach was found to be at least as effective at improving affinity while requiring fewer mutagenesis libraries than the traditional method. The resulting sequence data also highlight the potential for further reducing amino acid diversity for high affinity binding interactions.     

Active site structure and antigen binding properties of idiotypically cross-reactive anti-fluorescein monoclonal antibodies

This report includes complete VH and V kappa nucleotide and deduced amino acid sequences of idiotypically cross-reactive monoclonal anti-fluorescein antibodies that differed greater than 10(5)-fold in affinity. High affinity monoclonal antibody 4-4-20 and intermediate affinity antibodies 10-25, 5-14, 9-40, 12-40, and 3-24 utilized greater than or equal to 90% homologous VHIIIC germ-line genes. Extensive D segment length and sequence variability were observed; however, compensatory germ-line JH4 (4-4-20 and 3-24) or JH3 (10-25, 5-14, 9-40, and 12-40) sequence lengths resulted in H chain CDR3 + FR4 to be a constant 18 amino acids. In addition, each antibody and low affinity 3-13 rearranged greater than or equal to 96% homologous V kappa II genes to J kappa 1, except for 10-25 (J kappa 5) and 3-13 (J kappa 4). Resolved crystal structure of complexed fluorescein and 4-4-20 Fab fragments revealed residues HisL27d, TyrL32, ArgL34, SerL91, TrpL96, and TrpH33 acted as hapten contact residues. Antibodies 5-14, 9-40, 12-40, and 3-24 primary structures possessed identical contact residues as 4-4-20 except for the substitution of HisL34 for ArgL34. Thus, ArgL34 was implicated in the increased affinity of monoclonal antibody 4-4-20. Finally, it was difficult to correlate extensive H chain CDR3 residue heterogeneity directly with fluorescein binding and idiotypy.