The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cary, S. P.
Right arrow Articles by Silverman, G. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cary, S. P.
Right arrow Articles by Silverman, G. J.
The Journal of Immunology, 2000, 164: 4730-4741.
Copyright © 2000 by The American Association of Immunologists

Characterization of Superantigen-Induced Clonal Deletion with a Novel Clan III-Restricted Avian Monoclonal Antibody: Exploiting Evolutionary Distance to Create Antibodies Specific for a Conserved VH Region Surface1

Stephen P. Cary2, James Lee, Raymond Wagenknecht and Gregg J. Silverman3

The Sam and Rose Stein Institute for Research on Aging and the Theodore Gildred Cancer Center, Department of Medicine, University of California at San Diego, La Jolla, CA 92093


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Evolution of the Ab system has yielded three clans of VH region genes that are represented in almost every known higher species with an adaptive immune system. These clans are defined by sequence homologies primarily in highly conserved framework (FR) subdomains, which serve a scaffolding function maintaining the conformation of loops responsible for Ag binding. Structural analyses indicate that the VH FR1 and FR3 form a conserved composite exposed surface, which has been implicated in interactions with B cell superantigens. To directly investigate the expression of clan-defined supraclonal sets, we exploited the evolutionary distance of the chicken immune system and the selection power of phage display, to derive Abs diagnostic for clan III Ig. Using a specially tailored immunization and selection strategy, we created recombinant avian single chain Fv Abs specific for the clan III products, including those from the human VH3 family, and the analogous murine 7183, S107, J606, X24, and DNA4 families, and binding was competitive with natural B cell superantigens. The archetype, LJ-26, was demonstrated to recognize a clan-specific surface expressed in diverse mammalian, and also the Xenopus and chicken, immune systems. In flow-cytometric studies with LJ-26, we found that treatment of heterozygous T15i transgenic mice with a model B cell superantigen induced a clan III-restricted clonal deletion. These studies demonstrate the utility of a novel recombinant serologic reagent to study the composition of the B cell compartment and also the consequences of B cell superantigen exposure.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Understanding the evolutionary origins of lymphocyte Ag receptors, and discerning the molecular and natural selective pressures for diversification or conservation of the genes that encode them, are questions fundamental to the study of the adaptive immune system. Among the foundations of these studies are compilations of the sequences of Ig VH regions (1) that have led to an appreciation of their genetic organization. Based on their sharing of greater than 80% DNA sequence homology, the inherited gene segments encoding these VH regions have been organized into multigene families (2). Moreover, sequence analyses of human and murine Ab genes revealed that sets of VH gene families share highly conserved features, enabling their clustering into three major subgroups or clans (1, 3). From subsequent investigations of the Ab systems in diverse species, the progenitors of these VH clans have been postulated to have diverged even before the emergence of the mammalian radiation, as genes representative of the clans first appeared more than 300 million yr ago (4, 5, 6).

From efforts to dissect the functional capacities of these Ag receptors, V region sequences were shown to contain three noncontiguous linear intervals of greatest variability, which have been termed hypervariable regions or complementarity-determining regions (CDR)4 (7). Separating these CDR are intervals termed framework regions (FR) that are highly conserved among the members of a family. In crystallographic analyses that have elucidated the ß barrel structure of Abs, the CDR were found to represent loops that are juxtaposed to form the classic Ag binding site (reviewed in Refs. 8 and 9). By contrast, the FR subdomains fold into relatively rigid ß strands that maintain the overall Ig structure. However, these FR1 and FR3 subdomains were found to contain sequences that are distinct for each of the three clans. In fact, the nucleotide sequences of these FR subdomains are among the most highly conserved in mammals, with the FR1 and FR3 of clan III members displaying the greatest conservation (3, 10, 11, 12).

Competing theories have been presented to explain the maintenance of the clan-specific gene sequences across species and evolutionary boundaries. Tutter and Riblet (10, 11) compared inherited gene sequences and found that the frequencies of both silent and replacement mutations were significantly lower than expected, which was interpreted as indicative of selective conservation of clan III sequences at the nucleotide and not at the amino acid level. Presenting an alternative interpretation based on their own sequence analyses and molecular modeling of V regions, Kirkham and Schroeder (12) proposed that the FR subdomains did not solely provide structural and supportive scaffolding functions for the juxtaposed CDR sites. Instead, they demonstrated that the VH region FR1 and FR3 subdomains together form a separate conformational surface that is conserved and characteristic for each of the clans, which they postulated might represent an alternative ligand contact site (3, 12).

When originally presented, there was little more than circumstantial evidence to support this alternative binding site hypothesis. However, in recent years, certain microbial (13, 14, 15) and endogenous proteins (16, 17, 18) have been reported to have special properties, enabling direct interactions with these postulated Ig framework-associated alternative ligand-binding surfaces. Due to the obvious parallels with the activities of known T cell superantigens, many of these proteins have been postulated to represent B cell superantigens. We and others have considered the potential importance of the in vivo activities of a B cell superantigen (19), and speculated that interactions mediated through this Ig FR site might induce large scale alterations in the composition of the B cell or Ab repertoire. While this type of influence might affect all members of a species, possibly at an especially susceptible developmental stage, it is also possible that natural exposure may only occur occasionally in an individual host. However, despite their potential importance, except for our recent report (20), studies of these putative B cell superantigens have been limited to in vitro investigations.

The development of effective methods for monitoring of the consequences of in vivo B cell superantigen exposure poses special conceptual and technical challenges. One of the greatest hurdles is the lack of appropriate VH family-specific serologic markers. We suspect that the unavailability of this type of novel VH-targeted reagent is due to the conservation and ubiquitous expression of this same FR surface(s) in mammals, a state that must certainly have made these hosts immunologically tolerant to these determinants. Hence, we hypothesize that despite the obvious utility of this type of mAb, currently these reagents cannot be generated using available cellular technologies in commonly used experimental animals.

To surmount this postulated impediment, we sought to exploit the evolutionary distance of the chicken immune system. The idiosyncracies of the avian immune system are especially attractive for these studies (21, 22, 23, 24), as chicken Ab responses are formed from rearrangements of single VH and VL genes that are then highly modified by somatic mechanisms of hypermutation and gene conversion (reviewed in Refs. 25, 26). Therefore, to create these reagents, we first raised an Ab response to human monoclonal Ig in this avian host, and to isolate the mAbs of the required fine specificity, we harnessed the power of phage-display expression systems for in vitro clonal selection. By this approach, we were successful in isolating novel VH region-specific single chain V region (scFv) Abs specific for a conformational determinant that is completely restricted to the products of clan III genes. Moreover, in applications using diverse immunochemical formats, studies using these recombinant Abs demonstrated that clan III products are represented in the Ig repertoires of many terrestrial species. Further illustrating the utility of this novel highly specific clan III marker, we performed surveys of murine in vivo immune responses to a prototypic microbial B cell superantigen, which revealed the induction of a VH-restricted supraclonal defect. These studies provide a foundation for investigations into how a natural Ig-binding protein can mold the clonal composition within the B cell compartment.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Immunization and immunoassays of polyclonal IgY

A Leghorn chicken was immunized and boosted s.c. in the wing with human clan III Ig. For the immunization regimen, chickens received the native human IgM (V3–23/clan III V{kappa}1) 18/2 (27) and the human rIgM (V3–23/clan III V{kappa}2) Fab, 3-15 (28) (50 µg each), in PBS, pH 7.4 (PBS), emulsified in CFA (Difco Laboratories, Detroit, MI), with a boost after 2 wk of 3-15 and HEA (V3–30/clan III V{kappa}3) (50 µg each) in IFA. A second boost of 3-15 and 18/2 (50 µg each) in IFA was delivered 2 wk later. After 4 wk, induced Ab responses were demonstrated by comparisons of anti-Ig-binding activity of pre- and postimmunization IgY from egg yolk (data not shown), which was purified using the Eggstraction Kit (Promega, Madison, WI). Briefly, wells were coated with 3-15 at 5 µg/ml in PBS for 1 h at 37°C, and then blocked in 2% BSA in PBS for 1 h at 37°C. Equivalent concentrations of pre- and postimmunization IgY were added in serial dilutions in 1% BSA/PBS and incubated for 1 h at 37°C. To detect binding, peroxidase-labeled goat anti-IgY (Promega) was added in 1% BSA/PBS for 1 h at 37°C. Plates were washed as above, and tetramethylbenzidine substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) was added to wells. Plates were read at OD450–650 in a microplate reader (Bio-Rad, Hercules, CA).

Creation of a chicken scFv Ab phage-display library

Adapting a standard protocol (29), avian bone marrow and spleen were separately harvested into serum-free RPMI and dissociated into single cell mononuclear cell suspensions and then lysed in Tryzol (Life Technologies, Gaithersburg, MD). RNA was isolated by phenol/chloroform/isoamyl alcohol (24:25:1) extraction and precipitated in isopropanol at -20°C for 30 min. After centrifugation at 14,000 rpm in a microfuge, the RNA pellet was washed with cold 70% ethanol and recentrifuged. The pellet was air dried and resuspended in nuclease-free diethyl pyrocarbonate water. RNA was quantitated by OD260, and purity was determined by the 260:280 ratio. For reverse transcription, 20 µg of RNA was used with oligo(dT) primer according to the Superscript II kit (Life Technologies). Each of four PCR reactions for H chain and L chain genes used 1 µl of cDNA from the above reaction. The 100-µl PCR reaction also included 10x reaction buffer, 2.5 µM of dNTP, 1 µg each of either H chain primers (CSCVHo-FL3, GGTCAGTCCTCTAGATCTTCCGGCGGTGGTGGCAGCTCCGGTGGTGGCGGTTCCGCCGTGACGTTGGACGAG, and CSCG-B, CTGGCCGGCCTGGCCACTAGTGGAGGAGACGATGACTTCGGTCC) or L chain primers (CSCVk-F, GTGGCCCAGGCGGCCCTGACTCAGCCGTCCTCGGTGTC, and CKJo-B, GGAAGATCTAGAGGACTGACCTAGGACGGTCAGG). After a 2-min hot start at 96°C, 2.5 U of Taq polymerase was added to each reaction tube, and then thermal cycling was conducted in Perkin-Elmer 9600 (Applied Biosystems, Columbia, MD) for 30 cycles of 96°C for 30 s, 56°C for 15 s, 72°C for 90 s, followed by a final extension cycle of 72°C for 7 min. VH products (~430 bp) and VL products (~380 bp) were separately purified with a QIAquick kit (Qiagen, Chatsworth, CA). In 20 tubes for overlap PCR reactions, 100 ng of each of VH and VL PCR products was included with 2.5 µM of dNTP, and 1 µg each of nested sense primer (CSC-F, GAGGAGGAGGAGGAGGAGGTGGCCCAGGCGGCCCTGACTCAG) and antisense primer (CSC-B, GAGGAGGAGGAGGAGGAGGAGCTGGCCGGCCTGGCCACTAGTGGAGG), and nuclease-free water and the manufacturer’s PCR buffer in 100-µl reaction volumes. After a 2-min hot start at 96°C, 2.5 U of Taq polymerase was added to each reaction tube, with 30 cycles of 96°C for 30 s, 56°C for 15 s, 72°C for 2 min, followed by a final extension step at 72°C for 7 min. The products of these PCR reactions were combined, and the ~750-bp band was purified from a 2% agarose gel. Overlap product inserts and the pComb3H vector (30) (kind gift of Dr. Carlos Barbas III, Scripps Research Institute, La Jolla, CA) were prepared (20 µg of each) similarly by SfiI digestion for 5 h at 50°C, followed by ethanol precipitation, resuspension, and purification in a 2% agarose gel. Appropriate DNA bands were excised and electroeluted (Millipore, Bedford, MA). The ligation reaction included 700 ng of prepared single chain (sc) Fv library and 1400 ng of prepared pComb3H with 2 U of T4 DNA ligase (Life Technologies) incubated overnight at room temperature, which was later precipitated and resuspended in water before electroporation into electrocompetent ER2354 cells (NEB, Beverly, MA). By this approach, a library of >8.5 x 108 individual CFU was obtained.

Panning and binding studies of phage-display libraries

To obtain a phage-display form of the library, bacterial cultures were rescued overnight by infection with VCS M13 helper phage (Stratagene, La Jolla, CA) infection, and phage were precipitated in the morning in 4% PEG-8000 with 3% NaCl. For the first five rounds of panning, enzyme immunoassay microtiter wells were coated with human and/or mouse Ig at 5 µg/ml and blocked with 2% BSA/PBS, then phage were added alone or with a competitive inhibitor at 37°C for 2 h, and washed with PBS/0.05% Tween-20 (Table IGo). Bound phage were eluted with 0.1 M glycine-HCl, pH 2.2, and neutralized with 2 M Tris base. After each round of selection, fresh cultures of log phase ER2354 cells (NEB) were infected with the eluted phage, and, after library amplification, cultures were again rescued with helper phage. For the sixth round of selection, ~1012 phage were incubated with rocking in 750 µl of RPMI for 2 h at room temperature with ~108 murine mononuclear splenocytes from a naive 6-wk-old female BALB/c mouse. Afterward, cells were spun for 10 min at 4000 rpm in a microfuge, then washed thrice with 1 ml of PBS and respun, and then binding phage were eluted, as described above.


View this table:
[in this window]
[in a new window]
  		Table I. Selection strategy

 
To evaluate the outcome of each round of panning, binding activities of aliquots of isolated selected phage libraries were assessed in a phage-binding ELISA. For these studies, enzyme immunoassay microtiter wells were coated with purified monoclonal Ig, and blocked as before. Phage were then loaded in serial dilutions at equivalent titers, and binding was detected with HRP-labeled anti-M13 phage Ab (Pharmacia, Uppsala, Sweden), etc. To further define the phage library Ig fine binding specificity, the phage eluted from the fifth and sixth panning rounds were separately mixed with 10 µg/ml of a competitor clan III-specific Fab-binding protein, recombinant domain D of staphylococcal protein A (SpA) (28), and added to wells coated with the 3-15 Fab, and binding was detected as before.

Characterization of individual scFv Abs

The phage from the sixth round of selection were amplified in XL-1 Blue cells (Stratagene), and the plasmid library was purified (Qiagen). The scFv library was excised by SfiI digestion (NEB) for 3 h at 50°C, gel purified (Qiagen), and ligated into a similarly prepared pARA plasmid (gift of C. F. Barbas), a compatible bacterial expression vector employing the arabinose promoter that also fuses peptide tags to the scFv product. After transformation into XL-1 Blue cells, and selection on LB plates with 25 µg/ml of chloramphenicol, individual colonies were picked, and expanded in LB media with 25 µg/ml of chloramphenicol. At an OD600 of ~0.5, cultures were induced with 0.5% L-arabinose and grown overnight at 30°C. Protein was purified from cell pellets under native conditions over equilibrated Ni-NTA spin columns (Qiagen), and purified protein was eluted and immediately dialyzed in PBS, pH 8. In preparation for biotinylation, Abs were dialyzed against 0.25 M borate-buffered saline, pH 8.8, then reacted with long linker N-hydroxy-succinimide biotin (Sigma, St. Louis, MO) for 4 h at room temperature. The reaction was stopped with 1 M NH4Cl, and unreacted biotin was removed by dialysis against PBS for several days with four buffer changes. To evaluate individual scFv Abs for binding reactivity with a diverse panel of human and mouse monoclonal Ig, wells were coated and blocked as previously described, and equivalent amounts of scFv protein were loaded in serial dilutions. Binding was detected with a biotinylated anti-hemagglutinin reagent (gift of C. F. Barbas) that recognizes a pARA-encoded C-terminal epitope tag, with detection using HRP-streptavidin. Alternatively, in assays using directly biotinylated scFv Abs, binding was directly detected with HRP-streptavidin. To simplify comparisons with the Fab-binding activity of SpA, we used a chemically modified version (MSpA) that retains Fab specificity, but does not bind Fc{gamma} (31), which was biotinylated for detection of binding. For numeric comparisons, for each murine Ig the concentration values were determined from binding curves that provided an OD450–650 of 0.5 (i.e., stronger binding is associated with lower values).

For competition studies, a representative labeled scFv, LJ-26, or SpA was mixed with decreasing amounts of unlabeled LJ-26, SpA (Repligen, Cambridge, MA), or a human Fv-binding protein, termed pFv (18) (gift of J.-P. Bouvet, Hopital Broussais, Paris, France). The concentration of pFv was estimated based on silver stain of polyacrylamide gel studies and Western immunoblot analysis (not shown). Inhibition values were determined by interpolation against a standard curve of the labeled scFv without inhibitor.

The reactivities of select scFv clones were further characterized in immunoblots. Briefly, purified Ig from diverse species (Jackson ImmunoResearch, West Grove, PA) were separated on SDS-PAGE, under reducing and nonreducing conditions, in 4–12% Tris-glycine gels (Novex, La Jolla, CA), and later electrotransferred to Immobilon P membranes (Millipore). To assess binding, membranes were blocked in PBS/1% casein and blotted with scFv-biotin or a MSpA-biotin at 2 µg/ml, then washed in 0.05% Tween-20/borate-buffered saline. Reactivity was detected with HRP-streptavidin and chemiluminescent substrate (Amersham, Buckinghamshire, U.K.).

Microfluorometric studies

Samples of human PBMC containing greater than 95% monoclonal B cells from patients with chronic lymphocytic leukemia (CLL) (provided by T. J. Kipps, University of California, San Diego), and normal adult PBMC were stained with anti-CD5 PE (clone UCHT2), anti-CD19 APC (clone HIB19), anti-{kappa} FITC (G20-193), anti-{lambda} FITC (JDC-12), and LJ-26 biotin or biotin chicken scFv isotype control and streptavidin-peridinin chlorophyl protein. In studies of BALB/c, C57BL/6 or the AB29 (32), or T15i (33, 34) Ig transgenic mice, certain groups received neonatal treatment with endotoxin-free SpA, or hen egg lysozyme or saline, following a previously described treatment regimen (20). After the last treatment, bone marrow and splenic mononuclear cells were isolated as previously described (20) and evaluated freshly, or placed in overnight culture before analysis. Cells were treated then with Fc block, and stained with anti-IgM (R6-60.2), anti-B220 (RA3-6B2), anti-IgMa (DS-1), and LJ-26 biotin or biotinylated chicken scFv isotype control and streptavidin-peridinin chlorophyl protein. Unless indicated, Abs were obtained from PharMingen (La Jolla, CA). In certain studies, cells were also stained with a PE-labeled MSpA (20, 31). Data were acquired on a four-color FACSCalibur analytic flow cytometer (Becton Dickinson, Mountain View, CA) and analyzed with FloJo (Tree Star, San Carlos, CA) or CellQuest software (Becton Dickinson).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Selection of avian scFv Abs that bind clan III Ig

The immunization regimen was designed to enhance the response to determinants shared by human VH3 Abs, and the panning strat- egy was designed to positively select for clones with a cross-species clan III specificity, while clones specific for L chain and constant region determinants should be negatively selected (Table IGo). Hence, the avian scFv phage-display library was subjected to sequential rounds of selection against human and murine clan III Ig coated onto wells, and during the later rounds clan II Ig inhibitors were also mixed in solution with the phage libraries. For the final round, selection was against murine splenic mononuclear cells. The resulting libraries were subsequently evaluated for Ig-binding activity, which demonstrated that Ig-binding activity rose significantly after the fourth round of panning, and continued to increase for the fifth and sixth rounds (Fig. 1GoA).



View larger version (45K):
[in this window]
[in a new window]
  		FIGURE 1. Binding specificity of selected single chain avian Ab expression libraries. A, Illustrates that the binding activity of each selected scFv-phage library to purified monoclonal Ig increased with each sequential round of panning. In these comparative studies, microwells were coated with a murine (ms) J606/clan III IgG3, a murine 7183/clan III IgG1, a recombinant human (hu) VH3/clan III IgM Fab, or a human VH4/clan II IgM. After incubation with equivalent titers of 1011 phage, binding was detected with an anti-M13-HRP Ab-enzyme conjugate. In B, a fixed titer of the eluted libraries from panning rounds 5 and 6 was preincubated with, and without, a saturating concentration of a monomeric 11-kDa form of SpA, then later loaded to wells coated with the 3-15 VH3 IgM Fab. Residual binding was detected as described above. Significance was determined by the one-tailed unpaired Student t test: *, p = 0.029; **, p < 0.0001. Coincubation with control protein did not affect binding (not shown).

 
To assess whether these selected libraries contained an identifiable set of clones with a fine specificity for a clan III-restricted Fab surface, separate aliquots of phage from fifth and sixth rounds were mixed with a soluble competitor, an 11-kDa recombinant form of domain D of SpA (Fig. 1GoB). Studies with domain D were performed because this small monomeric protein expresses the parental clan III Fab-binding specificity (28), which targets a common and conserved conformational site formed by clan III-restricted residues primarily in the ß strands of the FR1 and FR3 subdomains (35). Domain D was responsible for significant inhibition of the binding to the clan III Ig of the fifth round scFv-phage library (p = 0.03, one-tailed Student t test), and inhibition was even greater for the sixth round scFv-phage library (p < 0.0001) (Fig. 1GoB). These data suggest that our panning strategy selected for a subpopulation of scFv-display phage with a fine specificity for a Fab site restricted to clan III Ig.

To characterize the reactivities of individual Ab clones, the sixth round library was transferred into a soluble scFv expression system, and the binding of purified scFv clones was tested by ELISA. Initial screening of 60 clones revealed a remarkable focusing of binding specificities, with representative results for 19 clones illustrated in Fig. 2GoA. Although the chicken was immunized with human IgM proteins, in these studies each of these soluble scFv also reacted with a murine 7183/clan III IgG1 and a murine J606/clan III IgG3, while these scFv were uniformly nonreactive with a clan II IgM. These data are consistent with specific interactions with cross-species-conserved VH region site(s).



View larger version (69K):
[in this window]
[in a new window]
  		FIGURE 2. Characterization of the Ig-binding specificities of representative soluble recombinant mAbs from the sixth round of selection. In A, within a set of soluble Ab clones, each individual clone is shown to recognize one or both of the murine clan III IgG, but did not bind a human IgM VH4/clan II. In B, the binding reactivities of a representative clone (LJ-26) were tested against a broad panel of human and mouse monoclonal Ig. This Ab displayed strong and near equivalent reactivity with diverse human and murine clan III Ig that was independent of VL region or C region usage. In contrast, there was no reactivity with Ig from the other clans. For these studies, a LJ-26 was directly biotinylated and loaded at a fixed dilution, and binding was detected by ELISA with labeled streptavidin.

 
A representative scFv Ab clone, LJ-26, was chosen for further characterization in assays with a diverse panel of human and murine Ig, in which LJ-26 bound to human Ig (3-15, 18/2, GLO, and MYE) from the VH3 gene family and to murine Ig from the S107 (EO6) and J606 VH families, all part of clan III. In contrast, LJ-26 did not recognize Ig from other VH clans, and binding also did not correlate with H chain isotype or VL usage (not shown) (Fig. 2GoB). Significantly, despite the presence of variations in the sequences of these clan III Abs, and the coassociation of diverse isotypes and L chains (Table IIGo), reactivity with LJ-26 was remarkably homogeneous.


View this table:
[in this window]
[in a new window]
  		Table II. VH regions and binding activity

 
Comparisons of the Fab-binding activities of LJ-26 and SpA

The relative reactivities of SpA and LJ-26 were compared in direct Fab-binding assays with murine IgM and IgG from clans I, II, and III, with results compiled in Table IIGo. Herein, the deduced protein sequences of the VH regions of these monoclonal Ig are grouped by family and clan, and aligned to the human germline V3–23 gene segment, which is believed to commonly encode for Abs in the human repertoire with the strongest Fab-mediated binding activity for SpA (14, 28). Similar to results from studies of SpA with monoclonal human Ig (13, 14, 28, 36) and murine Ig (37, 38), LJ-26 also did not exhibit detectable binding reactivity with any of the monoclonal murine Ig from clan I or II, which are significantly divergent from the V3–23 paragon, especially in key diagnostic positions in FR1 and FR3 (3). Most of the clan III IgM specificallyinteracted with both SpA and LJ-26, and strong reactivity was exhibited by the human VH3 and murine S107 and J606 family-encoded Ig. However, for certain mAbs (i.e., 7183-encoded Ig; 452s.11, 363p.16, and 452p.18), the binding activities of LJ-26 and SpA diverged, as they weakly bound SpA and were nonreactive with LJ-26. The greatest discordance was for several clan III IgG, including the 7183-encoded 452p.2 and MOPC21, which were all strongly reactive with LJ-26, while they bound SpA weakly or not at all. Furthermore, the lack of reactivity of MOPC21 with SpA has been correlated with a replacement mutation at position 57 and a natural variation of a serine at position 82a (38), which are reported to be critical residues for SpA binding (39, 40, 41). Hence, the strong reactivity of this avian Ab with MOPC21 may indicate that, while most of the contact surface is similar for SpA and LJ-26, these particular VH positions that are near the limits of the SpA contact surface (35) are not involved in the interaction with LJ-26. In general, while reactivity with SpA varied greatly between different clan III Ig, there was much less of a range of reactivities with LJ-26, which may suggest that the LJ-26 Ab recognizes a much less complex VH region surface.

LJ-26 competes with a bacterial B cell superantigen and an endogenous human Fab-binding protein for binding to clan III Ig

Competition studies were performed to directly compare the Fab-binding activity of LJ-26 with the natural clan III-restricted Ig-binding proteins, SpA, and the human gut-associated sialoprotein, pFv (18, 42). In the first type of competition immunoassay, a fixed concentration of labeled LJ-26 was mixed with different concentration of LJ-26, or an avian scFv isotype control, or pFv, or SpA, and then later incubated in wells coated with a human clan III IgM Fab. Binding of labeled LJ-26 to the human clan III Fab was efficiently competed by unlabeled LJ-26, SpA, and pFv (Fig. 3Go). Equivalent studies were also performed using labeled SpA, and each of these proteins significantly inhibited SpA binding (not shown). Together, these results suggest that LJ-26 binds at, or near, the same highly conserved clan III Fab surface that is responsible for the B cell superantigen activities of SpA and pFv. It is also important to appreciate that despite the fact that this scFv Ab has only a single binding site, in these studies the LJ-26 Ab displayed activity binding equal to or greater than the native SpA that has five domains capable of these Fab-binding interactions (28, 43).



View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 3. The avian LJ-26 Ab competes with SpA and the human gut-associated Fab-binding protein, protein Fv (pFv), for binding to a clan III Ig. For these studies, plates were coated with 3-15, a human VH3/clan III IgM Fab, and later serial dilutions of LJ-26, or SpA, or pFv or a control scFv were mixed with a fixed concentration of labeled LJ-26 Ab. Percent-inhibition values were derived by interpolation of the OD readings into a standard curve of labeled SpA or LJ-26 without inhibitors, as appropriate. Only the control scFv failed to compete for binding.

 
An avian recombinant Ab binds conformational VH region-associated determinant

To evaluate the structural requirements for VH region recognition, reactivity was assessed in immunoblot analysis with purified human and murine Ig. As shown in Fig. 4Go, both SpA and LJ-26 reacted only with the nonreduced clan III Ig, while reactivity was abolished under reducing conditions. Hence, the LJ-26 Ab, like the natural clan III VH-binding proteins, interacts with a conformational VH determinant that is abolished by these reducing conditions.



View larger version (40K):
[in this window]
[in a new window]
  		FIGURE 4. Western blot analysis of reactivity with human and murine Ig. Western analysis of a panel of purified Ig, in the following lanes: 1, polyclonal human adult IgG; 2, polyclonal human adult IgA; 3, polyclonal adult human IgM; 4, a human monoclonal VH3/clan III V{kappa}1, 18/2; and 5, a murine monoclonal IgM J558/clan I, M104E. In the top panels, proteins were separated under nonreducing conditions, while in the bottom panels, proteins were separated under reducing conditions. Replicates were developed with A, protein stain, while blots were reacted with B, the LJ-26 Ab, or C, MSpA, a form of SpA with only the Fab-binding specificity.

 
Based on the above described cross-species reactivity, we postulated that LJ-26 recognizes a clan III-associated determinant that has been conserved during the evolution of the adaptive immune system. To test this hypothesis, we ascertained the immunoblot reactivity of LJ-26 with a panel of purified Ig from diverse species. As illustrated in Fig. 5Go, LJ-26 specifically reacted with Ig from a wide range of placental mammals, including Rodentia (mouse and rat), Lagomorpha (rabbit), Carnivora (dog), Artiodactyla (pig, cattle, goat, and sheep), and Perissodactyla (horse). In addition, these limited surveys also demonstrated specific reactivity with an amphibian (Xenopus laevi), and unexpectedly also with an avian species, chicken, the host that was the source of the LJ-26 Ab. In general, reactivity with LJ-26 was comparable with the level of Fab-mediated SpA binding (Fig. 5GoB).



View larger version (51K):
[in this window]
[in a new window]
  		FIGURE 5. Western blot analysis of reactivity with Ig from diverse species. Western analysis was performed on purified Ig after nonreducing PAGE. In the following lanes: 1, polyclonal human adult IgM; 2, polyclonal human adult IgG F(ab')2; 3, rat IgG; 4, rabbit IgG; 5, swine IgG; 6, goat IgG; 7, horse IgG; 8, dog IgG; 9, bovine IgG; 10, sheep IgG; 11, frog Ig; 12, chicken Ig; 13, human IgG Fc; 14, monoclonal human VH4/clan II IgM. These replicates were developed with Coomassie stain (A), a form of SpA with only the Fab-binding specificity (MSpA) (B), LJ-26 (C), or an isotype control chicken scFv (D). Specificity was confirmed by lack of reactivity with IgG Fc or with the VH4/clan II IgM.

 
Flow-cytometric analyses of human and murine monoclonal B cells

To evaluate the reactivities of LJ-26 with B cell membrane-associated Ig, a series of microfluorometric assays were performed. In studies of PBMC from patients with chronic lymphocytic leukemia (CLL), which contain essentially monoclonal (i.e., >95%) CD5+CD19+ B cell populations, neither LJ-26 nor MSpA bound B cell CLL-expressing clan II Ig (e.g., COR) (Fig. 6GoA) or clan I Ig (not shown). In contrast, LJ-26 recognized most CLL that express clan III Ig (e.g., GOS and HEC), whereas MSpA bound fewer clan III-expressing CLL. These data document that LJ-26 recognizes a broad range of B cell-associated clan III Ig, displaying strong reactivity even with VH3-expressing B cells with low surface Ig levels.



View larger version (51K):
[in this window]
[in a new window]
  		FIGURE 6. Binding of the LJ-26 Ab is restricted to human B lymphocytes expressing membrane-associated clan III Ig. A, The peripheral B cells from patients with CLL were tested for binding the LJ-26 Ab, or a SpA reagent, or recombinant chicken Ab isotype control. These CLL samples contain essentially monoclonal B cell populations, as confirmed by CD19/CD5/{kappa}/{lambda} staining (not shown). Among the several samples studied, COR (VH4/clan II) expresses very high levels of surface Ig (data not shown), but was nonreactive with LJ-26 or the SpA reagent. For the VH3/clan III CLL, LJ-26 identified both GOS and HEC, while only HEC interacts with the SpA reagent. B, In CD19-gated human adult peripheral mononuclear cells, a similar proportion of {kappa}-staining B cells interacts with LJ-26 or the Fab binding site of SpA. C, In CD19-gated human adult peripheral mononuclear cells, a similar proportion of {lambda}-staining B cells interacts with LJ-26 or the Fab binding site of SpA.

 
In studies of PBMC from healthy adults, the reactivities of both MSpA (31) and LJ-26 were found to be restricted to a subset of CD19+ cells (i.e., B lymphocytes). Moreover, we found that the same ~25% of CD19+ gated mononuclear cells stained with MSpA or LJ-26 (Fig. 6Go, A and B). Of these CD19-positive populations, ~40% of the LJ-26+ and MSpA+ populations express {lambda} L chain, while ~60% of reactive populations express {kappa} L chains, which is the exact physiologic distribution of these L chains in the human immune system, indicating that these interactions are independent of B cell L chain usage.

The reactivity of the LJ-26 Ab was also evaluated in transgenic mice expressing defined VH and VL genes. In studies of AB29 mice, which have an expanded monoclonal B cell set expressing a human VH4/clan II µ-chain paired with a human V{kappa} L chain (32), as expected the splenic mononuclear cells bearing human IgM were not recognized by the LJ-26 Ab (Fig. 7GoA). However, in these mice, about one-third of the B220+ splenocytes instead express endogenous murine IgM, and LJ-26 recognized about 7% of these polyclonal splenic murine IgM-bearing B cells. These findings are consistent studies, in which 4–7% of C57BL/6 and 6–9% of BALB/c B220+/IgM+ splenocytes interact with the Fab binding site of SpA (20) (G. Silverman, unpublished observation).



View larger version (37K):
[in this window]
[in a new window]
  		FIGURE 7. Binding to B cells of Ig transgenic mice. A, Splenic mononuclear cells of AB29 mice include a large population of mature B220-positive cells (R1) expressing a transgene encoding surface membrane human IgM with a VH4/clan II-encoded region that is not recognized by the LJ-26 Ab. There is also a separate population of B220-positive lymphocytes expressing polyclonal endogenous IgM (R2), which includes a detectable proportion (~7%) of B cells recognized by LJ-26. B, Splenic mononuclear cells from T15i mice include a large population of B220-positive cells expressing the transgene-encoded IgMa (R3), of which most coexpress the S107/clan III/T15 rearrangement that is strongly reactive with the LJ-26 Ab.

 
We also studied T15i transgenic knockin mouse, which have B cells expressing a S107/clan III-VHT15 rearrangement associated with the IgMa allotype (33). In homozygous mice, we found that the great majority of these IgMa-reactive B cells were strongly reactive with LJ-26 (Fig. 7GoB) or the MSpA reagent (not shown), which is equivalent to results with a conventional VHT15-specific idiotypic marker (33). The remainder of the B220/CD45R-reactive cells instead express diverse endogenous VH rearrangements of the IgMb allele (33, 34), explaining the incomplete reactivity of all mature B cells with the clan III-specific reagents.

Investigations of a murine model of in vivo B cell superantigen-induced clonal defects

To test the hypothesis that natural B cell superantigens can affect the in vivo lymphocyte clonal composition, we investigated the consequences of exposure on heterozygous T15i x C57BL/6 F1 mice. Similar to the homozygous T15i mice, in these naive mice, most of the B cells that bear the IgMa allotype coexpress the H chain with the VHT15 rearrangement paired with diverse L chains (34); hence, there is not a significant population specific for any single conventional Ag. The remainder represent endogenous polyclonal B cells expressing the parental IgMb allotype, and hence, any fluctuations in the transgene-associated B cell set are measurable by use of VHT15-reactive and IgM allotype-specific markers.

For these studies, we treated the neonatal heterozygous mice with SpA or a control protein Ag, according to the same regimen recently used in studies demonstrating superantigen-induced supraclonal loss in BALB/c mice (20), and we evaluated the representation of peripheral B cells in mice sacrificed 18 h after introduction of the last dose. To ensure that a detected decrease in the representation of VHT15-expressing B cells is due to cellular deletion, and is not due to a reversible down-regulation of membrane-associated Ig, in certain studies these splenocytes were evaluated after 24 h of in vitro incubation in the absence of the immunogens. As illustrated in Fig. 8Go, compared with groups that received the control treatment, in which 32 ± 1.9% (mean ± SEM) of B cells were LJ-26 reactive, after SpA treatment only 3.7 ± 1.2% of B cells were LJ-26 reactive. Although due to gating differences, specific values were somewhat different, 41.1 ± 2.3% of the control-treated B cells expressed the transgene-associated IgMa allotype, while in SpA-treated mice the representation of IgMa-bearing B cells was reduced to 5.3 ± 1.4%, representing a highly significant 87% loss (p < 0.005, one-tailed Student t test). Also illustrated in Fig. 8Go, by gating on B220/CD45R-bearing cells, the concordance between IgMa and LJ-26 reactivity in these mice is directly demonstrated. Importantly, in the SpA-treated mice, the residual LJ-26-reactive B cells displayed lower levels of reactivity than detected in the control groups.



View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 8. Induction of a B cell clonal deletion in T15i heterozygous mice following exposure to SpA. (T15i x B6)F1 mice received eight i.p. doses of 100 µg of endotoxin-free SpA or hen egg lysozyme in 50 µl of saline every other day beginning within 24 h of birth (20 ). The day after the last dose, mice were sacrificed and splenocytes and bone marrow mononuclear cells directly evaluated, or cultured for 24 h in RPMI 1640 with 10% calf serum for 24 h before evaluation. In representative studies from overnight cultured splenocytes, the panels at left are from a mouse that received control (hen egg lysozyme) treatment, while the panels at right are from a mouse following exposure to SpA. The B cells expressing the VHT15 transgene represent a homogeneous population detected by costaining with B220 and the LJ-26 Ab (top row), or by costaining with B220 and the IgMa allotype-specific reagent (third row). The specificity of staining with LJ-26 is illustrated by the fewer than 1% of B220+ cells that interact with the isotype control avian scFv reagent (second row). Gating of mononuclear cells based on B220 reactivity (bottom row) demonstrated the concordance of strong reactivity of the LJ-26 with the transgene-expressing IgMa-reactive B cells. Treatment with SpA results in a greater than 90% deletion of the VHT15-expressing B cell population, which is comparable with recently reported findings in BALB/c based on MSpA binding (20 ). Data are representative of three multigroup studies.

 
In studies of freshly isolated cells, similar findings were obtained, with greater than 82% reductions in the representation of VHT15-expressing B cells, as detected by either LJ-26 or IgMa allotype reactivity (not shown). Notably, at this time point following exposure, there were no significant differences detected for the total number of splenocytes isolated from the two treatment groups, nor for the representation of mature B cells in the spleen. Therefore, our findings with the clan-specific marker were in close agreement with findings with the allotype-specific marker, documenting similar levels of B cell superantigen-induced postexposure losses of VHT15-expressing B cells in this peripheral compartment.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we describe a successful strategy that marries the special functional capacities of the avian immune system with the power of phage-display expression vectors for the efficient in vitro selection of Ab clones specific for highly conserved structural features of the mammalian immune system. By this approach, we have created unique Abs that are specific for a highly expressed VH region surface characteristic of clan III Abs.

To accomplish this goal, the avian immunization regimen consisted of sequential exposures with three monoclonal clan III Ig that derive from different human VH3 gene rearrangements, and that included L chains from diverse human V{kappa} genes. Subsequently, the genetic information for this Ab response was transferred into a suitable phagemid vector, and Abs with desirable specificities were isolated by a strategy in which only the initial round of panning used one of the original immunogens, under relatively low selection stringency. Later, to remove clones with irrelevant specificities, washing stringency was increased and selection was performed in the presence of soluble VH4/clan II IgM to prevent carryover of unwanted binders, including those to C region and L chain determinants. Within this strategy, a murine clan III Ig-selecting agent was next substituted and washing stringency was further increased. Both to reinforce selection for cross-species conserved determinants, and to ensure that isolated clones would be suitable for cell-staining studies, the final round was performed using selection upon viable murine splenic B cells.

Specificity analysis using a large number of human and murine samples documented that our strategy resulted in the recovery of clones reactive with the products of diverse genes from structurally related and highly expressed human and murine VH families. Reactivity was completely restricted to clan III Ig, with demonstrated complete nonreactivity with diverse clan I and clan II Ig. Predictably, reactivity with the representative LJ-26 Ab was much more frequent in the adult human peripheral B cell compartment than in the murine immune system, which is consistent with known species-specific VH gene family expression patterns (i.e., VH3/clan III is dominant in humans, while J558/clan I is dominant in mice).

From investigations of cross-species reactivity, we documented that the LJ-26 binding site, like the SpA binding site, is conserved on diverse mammalian species, activities generally consistent with reported Ab gene usage (discussed in Refs. 4, 5, 6, 12). We also demonstrated specific, but relatively weak, reactivity with sheep and bovine Ig, results seemingly in conflict with recent reports that these species do not have clan III genes (44, 45). However, clan III gene homologues have recently been cloned from sheep (John Reynolds, personal communication). We also found LJ-26 reactivity with frog Ig, which was predicted based on the deduced sequence of Xenopus V1 and related families (46). Although chicken VH genes are also clan III homologues, we did not expect to find such strong LJ-26 reactivity with chicken Ig. It is possible that this represents a neospecificity created by the in vitro combinatorial pairing of the avian VH and VL genes from cloning into the phagemid vector. However, because this Ab clone was rescued from a postimmunization response, we interpret this as more likely an indication that strict B cell immune tolerance is not maintained in this avian host.

The immunization and selection methods did not include SpA and pFv; therefore, we were somewhat surprised to discover that binding of the avian anti-clan III scFv was competitive with these clan III-specific natural B cell superantigens. Moreover, in several comparative assays of Abs with diverse VH regions, the binding specificities of SpA and LJ-26 were generally found to be quite similar. Akin to earlier reports of the specificity of superantigen binding (13, 14, 20, 38), LJ-26 was shown to recognize a conformational determinant that could be destroyed by reducing conditions. These findings may indicate that the clan III-specific VH surface responsible for the binding of B cell superantigens is highly accessible and perhaps dominant for immune recognition in a nontolerant host.

In studies of a panel of murine mAbs, reactivity of clan III Ig with LJ-26 was somewhat more common than interactions with the Fab binding site of SpA. Notably, LJ-26 recognized a product of the clan III/X24 family, while SpA did not. In addition, LJ-26 often had greater reactivity with clan III IgG, suggesting that somatic hypermutation may less commonly have an adverse effect upon recognition by the LJ-26 Ab. In microfluorometric assays of human CLL specimens, despite low surface Ig levels, more clan III CLL were also identified by LJ-26 than by SpA. We also found that in assays of polyclonal human peripheral B cells from healthy donors, LJ-26 staining displayed a closer direct correlation with the level of total surface Ig expression, providing a sharp diagonal. By comparison, although a very similar proportion of binders was recognized, there was much greater heterogeneity in the Fab-binding interactions of SpA. We speculate that the binding of the LJ-26 Ab may be less sensitive than SpA to local or remote mutational effects, perhaps because it interacts with a smaller surface that is less diversified in the products of different inherited and somatically mutated clan III genes. Based on the competitive inhibition studies, it is likely that, in part, these differences also reflect the higher binding affinity of the LJ-26 Ab.

Based on primary sequence correlation, we predict that the site on the surface of clan III VH region involved in the binding interaction of the LJ-26 Ab is a more limited surface than the contact site responsible for the binding of natural superantigens. Importantly, inherited polymorphisms and somatic variations at VH positions 57 and 82a do not correlate with differences in reactivity with the LJ-26 Ab, while these residues can greatly affect SpA-binding activity (39, 40, 41). Although the data are still too limited for meaningful mapping, it is likely that like SpA (35), the LJ-26 contact site is remote from the CDR loops involved in binding of conventional Ags, and may involve the VH ß strands of the FR1 and FR3 subdomains.

To directly evaluate the utility of the LJ-26 Ab for investigations of repertoire changes induced by B cell superantigen exposure, we used the T15i system generated by Taki et al. (33, 34), which we found was well suited to our goals. In part, this is because the expressed VHT15 transgene is paired with diverse endogenous L chains, which does not create a uniform population with a common conventional ligand-binding specificity. Hence, the effects induced by SpA exposure can only be due to unconventional VH region-mediated binding interactions (i.e., superantigen-VH framework). Most importantly, rearrangements of unmutated S107 genes commonly encode for B cell receptors with among the highest binding affinity for SpA, independent of heavy chain CDR3 or L chain usage (38), making it ideal for these investigations of the effects of a B cell superantigen on immune repertoire composition.

At a minimum, these studies have validated a strategy for the creation of novel reagents that are uniquely well suited to the study of the consequences of B cell superantigen exposure. Reiteration of this avian Ab-cloning approach is almost certain to provide reagents with complementary specificities to complete the overview of the expressed mammalian VH repertoire. Additional reagents for identifying further subdivisions of VH groupings, and their VL analogues should also be obtainable.

More importantly, we have advanced the hypothesis that Ig frameworks have functions beyond those of a passive scaffolding, i.e., rigid structural elements that position the CDR loops to create composite surfaces capable of binding diverse ligands. The VH surface created by the FR1/3 subdomains has been conserved even in primitive cartilaginous fish (reviewed in Ref. 47), which we believe is due to relationship(s) with novel environmental ligand(s), or perhaps due to their interactions with special Ig adapter molecules conveying functions that reiterate themes first appreciated for Fc regions. These studies have also provided the first evidence that in the suitable host, these conserved surfaces can be recognized by the immune system. In the future, serologic tools akin to LJ-26 should advance our understanding of the impact of poorly understood infectious, inflammatory, and autoimmune disease processes upon the B cell compartment.


   Acknowledgments
 
We acknowledge the invaluable technical assistance contributed by Denise Dwyer-Pipkin, Rowena Aguilar-Sino, and Linda Luo. We appreciate the Ab samples provided by Dr. Tony Marion, the display vectors and protocols provided by Dr. Carlos Barbas, the CLL samples provided by Drs. T. J. Kipps and Laura Rassenti, and the AB29 transgenic mice provided by Dr. Helen Tighe. The T15i trangenic mice from Dr. Klaus Rajewsky were kindly transferred to us by Drs. Betty Diamond and Christine Grimaldi. We also thank Drs. Enrico Stura and Adam Corper for helpful discussions on the structural basis for these interactions, and Dr. Carl Goodyear for assistance in cross-species analyses.


   Footnotes
 
1 This work was supported by National Institutes of Health Grants R01-AI40305 and 5P60-AR40770. G.J.S. was the recipient of a Career Development Award from the National Institute of Allergy and Infectious Diseases (K02-AI01378) and a Biomedical Sciences Award from the Arthritis Foundation. Back

2 Current address: Department of Biological Chemistry, University of Michigan Medical School, Ann Arbor, MI 48109. Back

3 Address correspondence and reprint requests to Dr. Gregg J. Silverman, Department of Medicine-0663, University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0663. E-mail address: Back

4 Abbreviations used in this paper: CDR, complementarity-determining region; CLL, chronic lymphocytic leukemia; FR, framework region; HEL, hen egg lysozyme; MSpA, chemically modified SpA that retains Fab-binding activity; pFv, protein Fv; scFv, single chain Fv; SpA, staphylococcal protein A. Back

Received for publication October 5, 1999. Accepted for publication February 22, 2000.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Kabat, E. A., T. T. Wu, H. M. Perry, K. S. Gottesman, C. Foeller. 1991. Sequences of Proteins of Immunological Interest National Institutes of Health, Bethesda, MD.
  2. Brodeur, P. H., R. Riblet. 1984. The immunoglobulin heavy chain variable region (IgH-V) locus in the mouse. I. One hundred IgH-V genes comprise seven families of homologous genes. Eur. J. Immunol. 14:922.[Medline]
  3. Kirkham, P. M., F. Mortari, J. A. Newton, Jr H. W. Schroeder. 1992. Immunoglobulin VH clan and family identity predicts variable domain structure and may influence antigen binding. EMBO J. 11:603.[Medline]
  4. Jr Schroeder, H. W., J. L. Hillson, R. M. Perlmutter. 1990. Structure and evolution of mammalian VH families. Int. Immunol. 2:41.[Abstract/Free Full Text]
  5. Marchalonis, J. J., S. F. Schluter, R. M. Bernstein, S. Shen, A. B. Edmundson. 1998. Phylogenetic emergence and molecular evolution of the immunoglobulin family. Adv. Immunol. 70:417.[Medline]
  6. Andersson, E., T. Matsunaga. 1995. Evolution of immunoglobulin heavy chain variable region genes: a VH family can last for 150–200 million years or longer. Immunogenetics 41:18.[Medline]
  7. Kabat, E. A., T. T. Wu, H. Bilofsky. 1977. Unusual distributions of amino acids in complementarity-determining (hypervariable) segments of heavy and light chains of immunoglobulins and their possible roles in specificity of antibody-combining sites. J. Biol. Chem. 252:6609.[Free Full Text]
  8. Wilson, I. A., R. L. Stanfield. 1993. Antibody-antigen interactions. Curr. Biol. 4:857.
  9. MacCallum, R. M., A. C. Martin, J. M. Thornton. 1996. Antibody-antigen interactions: contact analysis and binding site topography. J. Mol. Biol. 262:732.[Medline]
  10. Tutter, A., R. Riblet. 1989. Conservation of an immunoglobulin variable-region gene family indicates a specific, noncoding function. Proc. Natl. Acad. Sci. USA 86:7460.[Abstract/Free Full Text]
  11. Tutter, A., R. Riblet. 1988. Selective and neutral evolution in the murine IgH-V locus. Curr. Top. Microbiol. Immunol. 137:107.[Medline]
  12. Kirkham, P. M., Jr H. W. Schroeder. 1994. Antibody structure and the evolution of immunoglobulin V gene segments. Semin. Immunol. 6:347.[Medline]
  13. Sasso, E. H., G. J. Silverman, M. Mannik. 1989. Human IgM molecules that bind staphylococcal protein A contain VHIII H chains. J. Immunol. 142:2778.[Abstract]
  14. Sasano, M., D. R. Burton, G. J. Silverman. 1993. Molecular selection of human antibodies with an unconventional bacterial B cell superantigen. J. Immunol. 151:5822.[Abstract]
  15. Berberian, L., L. Goodglick, T. J. Kipps, J. Braun. 1993. Immunoglobulin VH3 gene products: natural ligands for HIV gp120. Science 261:1588.[Abstract/Free Full Text]
  16. Lenert, P., D. Kroon, H. Spiegelberg, E. S. Golub, M. Zanetti. 1990. Human CD4 binds immunoglobulin. Science 248:1639.[Abstract/Free Full Text]
  17. Pospisil, R., G. O. Young-Cooper, R. G. Mage. 1995. Preferential expansion and survival of B lymphocytes based on VH framework 1 and framework 3 expression: "positive" selection in appendix of normal and VH-mutant rabbits. Proc. Natl. Acad. Sci. USA 92:696.
  18. Silverman, G. J., P. Roben, J.-P. Bouvet, M. Sasano. 1995. Superantigen properties of a human sialoprotein involved in gut-associated immunity. J. Clin. Invest. 96:417.
  19. Silverman, G. J.. 1998. B cell superantigens: possible roles in immunodeficiency and autoimmunity. Semin. Immunol. 10:43.[Medline]
  20. Silverman, G. J., J. V. Nayak, K. Warnatz, S. Cary, H. Tighe, V. E. Curtiss. 1998. The dual phases of the response to neonatal exposure to a VH family-restricted staphylococcal B-cell superantigen. J. Immunol. 161:5720.[Abstract/Free Full Text]
  21. Yamanaka, H. I., T. Inoue, O. Ikeda-Tanaka. 1996. Chicken monoclonal antibody isolated by a phage display system. J. Immunol. 157:1156.[Abstract]
  22. Davies, E. L., J. S. Smith, C. R. Birkett, J. M. Manser, D. V. Anderson-Dear, J. R. Young. 1995. Selection of specific phage-display antibodies using libraries derived from chicken immunoglobulin genes. J. Immunol. Methods 186:125.[Medline]
  23. Michael, N., M. A. Accavitti, E. Masteller, C. B. Thompson. 1998. The antigen-binding characteristics of mAbs derived from in vivo priming of avian B cells. Proc. Natl. Acad. Sci. USA 95:1166.[Abstract/Free Full Text]
  24. Reynaud, C. A., A. Dahan, V. Anquez, J. C. Weill. 1989. Somatic hyperconversion diversifies the single VH gene of the chicken with a high incidence in the D region. Cell 59:171.[Medline]
  25. Weill, J. C., C. A. Reynaud. 1987. The chicken B cell compartment. Science 238:1094.[Abstract/Free Full Text]
  26. McCormack, W. T., L. W. Tjoelker, C. B. Thompson. 1993. Immunoglobulin gene diversification by gene conversion. Prog. Nucleic Acid Res. Mol. Biol. 45:27.[Medline]
  27. Dersimonian, H., R. S. Schwartz, K. J. Barrett, B. D. Stollar. 1987. Relationship of human variable region heavy chain germ-line genes to genes encoding anti-DNA antibodies. J. Immunol. 139:486.
  28. Roben, P., A. Salem, G. J. Silverman. 1995. VH3 antibodies bind domain D of staphylococcal protein A. J. Immunol. 154:6437.[Abstract]
  29. Andris-Widhof, J., P. Steinberger, G. J. Silverman, and C. F. Barbas. 2000. Library construction. In Phage Display: A Laboratory Manual. Cold Spring Harbor Laboratories Press, New York, in press.
  30. Yang, W. P., K. Green, S. Pinz-Sweeney, A. T. Briones, D. R. Burton, III C. F. Barbas. 1995. CDR walking mutagenesis for the affinity maturation of a potent human anti-HIV-1 antibody into the picomolar range. J. Mol. Biol. 254:392.[Medline]
  31. Silverman, G. J., M. Sasano, S. B. Wormsley. 1993. Age-associated changes in binding of human B lymphocytes to a VH3-restricted unconventional bacterial antigen. J. Immunol. 151:5840.[Abstract]
  32. Tighe, H., P. Heaphy, S. Baird, W. O. Weigle, D. A. Carson. 1995. Human immunoglobulin (IgG) induced deletion of IgM rheumatoid factor B cells in transgenic mice. J. Exp. Med. 181:599.[Abstract/Free Full Text]
  33. Taki, S., M. Meiering, K. Rajewsky. 1993. Targeted insertion of a variable region gene into the immunoglobulin heavy chain locus. Science 262:1268.[Abstract/Free Full Text]
  34. Taki, S., F. Schwenk, K. Rajewsky. 1995. Rearrangement of upstream DH and VH genes to a rearranged immunoglobulin variable region gene inserted into the DQ52-JH region of the immunoglobulin heavy chain locus. Eur. J. Immunol. 25:1888.[Medline]
  35. Graille, M., E. A. Stura, A. L. Corper, B. J. Sutton, M. J. Taussig, J.-B. Charbonnier, and G. J. Silverman. 2000. Crystal structure of a Staphylococcus aureus protein A domain complexed with the Fab fragment of a human IgM antibody: structural basis for recognition of B-cell receptors and superantigen activity. Proc. Natl. Acad. Sci. USA, In press.
  36. Sasso, E. H., G. J. Silverman, M. Mannik. 1991. Human IgA and IgG F(ab')2 that bind to staphylococcal protein A belong to the VHIII subgroup. J. Immunol. 147:1877.[Abstract]
  37. Seppala, I., M. Kaartinen, S. Ibrahim, O. Makela. 1990. Mouse Ig coded by VH families S107 or J606 bind to protein A. J. Immunol. 145:2989.[Abstract]
  38. Cary, S., M. R. Krishnan, T. Marion, G. J. Silverman. 1999. The murine clan VH III related 7183, J606 and S107 and DNA4 families commonly encode for binding to a bacterial B cell superantigen. Mol. Immunol. 36:769.[Medline]
  39. Randen, I., K. N. Potter, Y. Li, K. M. Thompson, V. Pascual, O. Forre, J. B. Natvig, J. D. Capra. 1993. Complementarity-determining region 2 is implicated in the binding of staphylococcal protein A to human immunoglobulin VHIII variable regions. Eur. J. Immunol. 23:2682.[Medline]
  40. Hakoda, M., S. Hayashimoto, H. Yamanaka, C. Terai, N. Kamatani, S. Kashiwazaki. 1994. Molecular basis for the interaction between human IgM and staphylococcal protein A. Clin. Immunol. Immunopathol. 72:394.[Medline]
  41. Zhang, M., A. Majid, P. Bardwell, C. Vee, A. Davidson. 1998. Rheumatoid factor specificity of a VH3-encoded antibody is dependent on the heavy chain CDR3 region and is independent of protein A binding. J. Immunol. 161:2284.[Abstract/Free Full Text]
  42. Silverman, G. J., R. Pires, J. P. Bouvet. 1996. An endogenous sialoprotein and a bacterial B cell superantigen compete in their VH family-specific binding interactions with human Igs. J. Immunol. 157:4496.[Abstract]
  43. Jansson, B., M. Uhlen, P. A. Nygren. 1998. All individual domains of staphylococcal protein A show Fab binding. FEMS Immunol. Med. Microbiol. 20:69.[Medline]
  44. Saini, S. S., W. R. Hein, A. Kaushik. 1997. A single predominantly expressed polymorphic immunoglobulin VH gene family, related to mammalian group, I, clan, II, is identified in cattle. Mol. Immunol. 34:641.[Medline]
  45. Dufour, V., S. Malinge, F. Nau. 1996. The sheep Ig variable region repertoire consists of a single VH family. J. Immunol. 156:2163.[Abstract]
  46. Mussmann, R., M. Courtet, L. Du Pasquier. 1998. Development of the early B cell population in Xenopus. Eur. J. Immunol. 28:2947.[Medline]
  47. Marchalonis, J. J., R. M. Bernstein, S. X. Shen, S. F. Schluter. 1996. Emergence of the immunoglobulin family: conservation in protein sequence and plasticity in gene organization. Glycobiology 6:657.[Free Full Text]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
C. S. Goodyear, F. Sugiyama, and G. J. Silverman
Temporal and Dose-Dependent Relationships between In Vivo B Cell Receptor-Targeted Proliferation and Deletion-Induced by a Microbial B Cell Toxin
J. Immunol., February 15, 2006; 176(4): 2262 - 2271.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
C. S. Goodyear and G. J. Silverman
Staphylococcal toxin induced preferential and prolonged in vivo deletion of innate-like B lymphocytes
PNAS, August 3, 2004; 101(31): 11392 - 11397.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
A. J. Cappione, A. E. Pugh-Bernard, J. H. Anolik, and I. Sanz
Lupus IgG VH4.34 Antibodies Bind to a 220-kDa Glycoform of CD45/B220 on the Surface of Human B Lymphocytes
J. Immunol., April 1, 2004; 172(7): 4298 - 4307.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
C. S. Goodyear, M. Narita, and G. J. Silverman
In Vivo VL-Targeted Activation-Induced Apoptotic Supraclonal Deletion by a Microbial B Cell Toxin
J. Immunol., March 1, 2004; 172(5): 2870 - 2877.
[Abstract]