FASEB J.
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
Services
Right arrow Email this article to a friend
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
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by ROBEY, I. F.
Right arrow Articles by MARCHALONIS, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by ROBEY, I. F.
Right arrow Articles by MARCHALONIS, J. J.
(The FASEB Journal. 2002;16:642-652.)
© 2002 FASEB

Specificity mapping of human anti-T cell receptor monoclonal natural antibodies: defining the properties of epitope recognition promiscuity

IAN F. ROBEY, ALLEN B. EDMUNDSON*, SAMUEL F. SCHLUTER, DAVID E. YOCUM{dagger} and JOHN J. MARCHALONIS1

Department of Microbiology and Immunology; and the
{dagger} Arizona Arthritis Center, College of Medicine, University of Arizona, Tucson, Arizona, USA; and
* Oklahoma Medical Research Foundation, Oklahoma City, Oklahoma, USA

1Correspondence: Department of Microbiology and Immunology; and the {dagger}Arizona Arthritis Center, College of Medicine, University of Arizona, Tucson, AZ 85724, USA.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The classical concept of antibody binding is defined as an exclusive and high-affinity interaction with one epitope. The emerging reality about antibody combing sites, however, is that some can bind unrelated determinants. The studies presented here define this quality as epitope recognition promiscuity by analyzing the capacity of monoclonal human autoantibodies to bind sets of overlapping peptides duplicating the complete structures of T cell receptor (TCR) {alpha} and ß chains and immunoglobulin {lambda} chain. We assessed the binding of these monoclonal antibodies (mAbs) to a set of homologous peptides corresponding to the CDR1 segments of human Vß gene products, a major epitope used in the selection of the antibodies. We present data on the binding characteristics of four human mAbs selected for the ability to bind TCR epitopes. These mAbs are IgM molecules with VH and VL sequences in germline configuration, but have diverse VH CDR3 regions. These studies aim to characterize the property of epitope promiscuity and show that the relationship between the binding site and its epitope is a complex interaction and unpredictable from antigen sequence alone. Our results support the conclusion that epitope recognition promiscuity is a genuine feature of antibody and TCR recognition.—Robey, I. F., Edmundson, A. B., Schluter, S. F., Yocum, D. E., Marchalonis, J. J. Specificity mapping of human anti-T cell receptor monoclonal natural antibodies: defining the properties of epitope recognition promiscuity.


Key Words: antigen binding site • polyreactivity • mAbs • affinity


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
IMMUNOGLOBULINS (IGS) SERVE as the principal molecular components of the humoral immune system. Comprised of protein structures that form multivalent binding sites, they function as a first line of defense, coupled with long-term protection, against tissue damage and infection through the neutralization of toxic and pathogenic determinants. A remarkable feature of Igs is that each possesses an antigen binding site that can discriminate between discrete molecular structures (1) . The existence of such a comprehensive binding motif network, capable of responding to millions of diverse antigens, depends on a highly complex combinatorial immune system providing an extreme level of variability among the Ig pool (2) .

Studies of antibody specificity have yielded many discoveries about the nature of antigen binding. The variable (V) region sequences determine what epitope and how tightly the antibody will bind, but to what extent cannot be predicted. The basic model for the antibody-to-antigen interaction consists of a specific binding to a single epitope at relatively high affinity. This notion was deduced from the clonal selection theory proposed by Burnet in the 1950s (3) . Subsequent investigations have revealed that this concept is too simplified to represent all antibody/antigen interactions. New terms were required to account for antibody binding to noncognate antigen. Some definitions can be understood easily. ‘Cross-reactivity’, for example, describes an interaction between an antibody and epitopes that are antigenically related. Binding affinity differences for related structures depend on the spatial orientation of functional groups of the antigenic determinant (4) . This can be seen at the protein and peptide level. ‘Polyreactivity’ describes a different but not highly unusual quality of antibody binding to many large, complex, and commonly occurring molecular structures with repeating antigenic motifs. Typically these antibodies are IgM molecules that exist in the serum before immune priming. Polyreactive antibodies are generally known to bind these determinants with low affinity (5 , 6) .

Epitope recognition promiscuity differs from polyreactivity in that the antibody can specifically recognize distinct epitopes at a different or similar affinity, and these determinants compete for the same binding site. Promiscuous antibodies are restricted in binding to a limited set of epitopes and are not thought to bind in a nonspecific manner (7 , 8) . This binding property is observed in both IgM and IgG molecules (9 10 11) and exists in antibodies from the most primitive combinatorial immune systems (8 , 12 , 13) . There is limited speculation, however, on the functional purpose of epitope recognition promiscuity.

A distinguishable property observed in natural antibodies is the variability of the amino acid sequence in the CDR3 V region heavy chain. These unique sequence segments are products of the imprecise joining events between the V, diversity (D), and joining (J) regions of the heavy chain and subsequent N region additions that occur during V gene rearrangement. It is hypothesized that heavy chain CDR3 diversity is the critical element that determines whether an antibody will be polyreactive epitope promiscuous or monospecific for antigen (14 15 16 17 18 19 20) .

From studies determining the specificities for defined synthetic peptides of monoclonal natural antibodies selected for the ability to bind to the same human T cell receptor (TCR) antigens, we report that epitope recognition promiscuity is a genuine property distinguishable from the occurrence of polyreactivity. Our findings demonstrate that natural antibodies with diverse heavy chain CDR3 segments but otherwise in germline sequence are able to bind specifically to sequentially unrelated peptide epitopes, but not to test proteins typically recognized by polyreactive antibodies. We further show that epitope recognition promiscuity can be expressed in various degrees and affinities to different TCR epitopes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Antigens and antibodies
The following synthetic peptides were used for fine specificity mapping and titration analysis of the anti-TCR mAbs: 1) 22 TCR-ß peptides corresponding to the complete Vß 8.1 YT35 TCR (21) , 2) 19 TCR-{alpha} peptides corresponding to the complete V{alpha} 1 pY14 TCR (22) , 3) 20 Mcg peptides corresponding to the complete Mcg {lambda} light chain sequence (23) , 4) 24 Vß TCR CDR1 homologue peptides, and 5) 6 Vß TCR FR3 homologue peptides (24) . The panel of TCR-ß, TCR-{alpha}, and Mcg 16-mer peptides overlap each other sequentially by five amino acids from the amino (left) to the carboxyl (right) terminus of the complete protein (25) (Table 1 , Table 2 , Table 3 ). The Vß TCR CDR1 and FR3 16- and 17-mer homologs represent CDR1 (and partial FR2) (Table 4) and FR3 segments (Table 5) , reading from the amino (left) to the carboxyl (right) terminus, corresponding to 25 Vß gene products. All peptides were synthesized to 95% purity by Chiron Mimetopes® (San Diego, CA) or the University of Arizona Biotechnology Center (Tucson, AZ). Whole protein antigens were fetuin, bovine serum albumin (BSA), ovalbumin, thyroglobulin (Sigma, St. Louis, MO), and pooled human polyclonal IgG Gammagard® (Baxter, Deerfield, IL) as polyreactivity controls.


View this table:
[in this window]
[in a new window]
  		Table 1. TCR ß peptides


View this table:
[in this window]
[in a new window]
  		Table 2. TCR {alpha} peptides


View this table:
[in this window]
[in a new window]
  		Table 3. Mcg Ig {lambda} peptides


View this table:
[in this window]
[in a new window]
  		Table 4.


View this table:
[in this window]
[in a new window]
  		Table 5.

Antibodies used for these assays (OR1, OR2, OR5, and Syn 2H-11) were human mAbs selected from hybridoma limit dilution assays for their ability to bind a 16 amino acid peptide corresponding to the complete CDR1 and part of the FR2 segment of the Vß 8.1 TCR and a recombinant single-chain TCR containing the complete VJ{alpha} and VDJß of the JURKAT T cell (26) . Monoclonal antibodies (mAbs) OR1, OR2, and OR5 were derived from the peripheral blood lymphocytes of a patient with rheumatoid arthritis (RA) and mAb Syn 2H-11 was derived from the synovial tissue lymphocytes of a patient with RA (27) . A plasma sample from an RA individual was used as a positive control for rheumatoid factor on ELISAs to test reactivity to pooled human polyclonal IgG.

Enzyme-linked immunosorbent assay
Cell culture supernatants were tested by enzyme-linked immunosorbent assay (ELISA). Assay plates (NuncTMBrand Products, Kamstrupvej, Denmark) were incubated overnight at 4°C with 100 µl of 10 µg/ml whole protein antigen (with the exception of scTCR used for the inhibition experiments, which was coated on plates at 1.25 µg/ml) or dried down at 37°C with 5 µg/ml peptide in 0.2 M sodium carbonate buffer, pH 9.6. Plates were blocked for 1 h at room temperature with SuperBlock® blocking buffer in PBS (Pierce, Rockford, IL). Supernatants were applied in twofold serial dilutions starting at 1/10 in PBS-Tween (Tween at 0.5 ml/L in PBS) and incubated at room temperature for 1 h. After washing, rabbit anti-human IgM heavy chain secondary antibody (Dako, Glostrup, Denmark) conjugated with horseradish peroxidase was applied in 1/3000 dilution in PBS-Tween for 1 h at room temperature. Plates were developed with 30 mM 2,2'-azino-bis(3-ethylbenz-thiazoline-6-sulfonic acid)(ABTS) reagent in citrate buffer, pH 4.0. Color changes on the plates were measured on a plate reader (Titertek Multiskan®) at wavelength 405 nm. Uncoated wells were left on the same plates during the procedure to compare supernatant binding to plastic with antigen.

Determining antibody concentration
Antibody concentration was determined by capture ELISA. Assay plates were coated with 10 µg/ml of anti-human {lambda} light chain antibody (Sigma). Known dilutions of human IgM{lambda} myeloma protein (The Binding Site, Birmingham, UK) were used as the control antigen to plot a standard curve. The slope of the best-fit curve was used to determine the concentration of unknown antibody.

Inhibition assay
Before inhibition, assay wells in ELISA plates were preblocked overnight (~ 16 h) at 4°C. Inhibiting antigen was diluted in PBS-Tween with 0.1% BSA and added to the blocked wells. The starting concentrations of inhibiting antigen (50 µg/ml of peptide and 25 µg/ml of scTCR) were carried out twofold in six more serial dilutions. The TCR-ß1 peptide (Table 1) was used as a specificity control for the soluble inhibiting TCR antigens. One set of wells was treated with assay diluent without soluble antigen for maximum antibody binding measurement. Antibody was added at ~2 µg/ml to the wells; the plates were covered and incubated on a rotator overnight at 4°C. Inhibition reactions were applied to antigen-coated plates for ELISA (see above).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Anti-TCR mAbs demonstrate differential binding profiles on TCR and Ig light chain peptides
We previously reported the generation and characterization of seven clones from RA patients, each with different variable region heavy and light chain sequences but selected on the same TCR antigens (27) . Four of these mAbs—OR1, OR2, OR5, and Syn 2H-11—were tested for binding to three sets of sequentially overlapping peptides ~16 amino acids in length, corresponding directly to the complete pY14 {alpha} TCR, YT35 ß TCR, and Mcg Ig {lambda} chain sequence. The anti-TCR mAbs were tested on peptides corresponding to antibody light chain because the Mcg {lambda} sequence is structurally homologous to the TCR-ß sequence, and thus helpful in assessing the degree of cross-reactivity between these peptides. A minimum of three ELISAs were conducted (at separate intervals) for each anti-TCR mAb on the different peptide groups. The reciprocals of the supernatant dilutions at which the absorbance measurement (OD 405) was 0.5 units were compiled and expressed as a geometric mean titer to designate relative antibody reactivity. This method allows for comparison of data from different assay dates, but excludes tenuous and false-positive assay results.

As indicated in Fig. 1 , relative antibody binding profiles differed significantly on the TCR-ß peptides. Anti-TCR mAb OR1 bound exclusively to ß3. This is the CDR1/FR2 peptide on which all the anti-TCR mAbs were selected (27) . OR1 tested on the TCR-{alpha} peptides reacted only with {alpha}3 and {alpha}4 (Fig. 2 ). It did not bind with any of the Mcg Ig {lambda} peptides (Fig. 3 ). The mAb OR2 illustrates a very different quality among the anti-TCR mAbs. When examined on the TCR-ß peptides, OR2 reacted substantially with five peptides besides ß3. These peptides were ß8, ß11, ß14, ß15, and ß20. OR2 bound to peptides that corresponded to the FR3 (ß8) region of the Vß TCR, parts of the J and constant (C) region (ß11), and at least two different segments of the C region (ß14/ß15 and ß20). There is little if any comparative sequence homology between these peptides. OR2 appeared to demonstrate the same kind of polyreactive qualities on the TCR-{alpha} peptides (Fig. 2) . OR2 bound similarly with OR1 to {alpha}3 and {alpha}4 and the {alpha}5 and {alpha}6 peptides. OR2 demonstrated reactivity to the {alpha}9 peptide, which is part of the V region CDR3 sequence, and TCR-{alpha} peptides 12 through 17, all of which correspond to the TCR-{alpha} constant region sequence. OR2 bound to a set of Ig {lambda} Mcg peptides corresponding to the CDR1 region (Mcg 4 and 5), the FR3 and CDR3 region (Mcg 8), the J region (Mcg 10), and the constant region (Mcg 13, 15–17, 20) of the {lambda} chain (Fig. 3) . Anti-TCR mAbs OR5 and Syn 2H-11 demonstrated more restrictive binding patterns to the test peptides. Besides the ß3 peptide, OR5 and Syn 2H-11 bound to another ß peptide corresponding to the constant region. OR5 bound to the ß17 peptide and Syn 2H-11 bound to the ß22 peptide (Fig. 1) . On the TCR-{alpha} peptides, OR5 and Syn 2H-11 reacted to {alpha}3 and {alpha}4, as observed with OR1 and OR2. OR5 bound to the {alpha}16 peptide of the TCR-{alpha} constant region, but Syn 2H-11 reacted with no other {alpha} peptides (Fig. 2) . OR5 and Syn 2H-11 bound to Mcg CDR1 region peptides 3 and 4 (similar to OR2) and two other Mcg peptides each. OR5 reacted to Mcg 8 and 17 and Syn 2H-11 reacted to Mcg 13 and 17 (Fig. 3) .



View larger version (20K):
[in this window]
[in a new window]
  		Figure 1. Anti-TCR mAbs binding to TCR-ß peptides. Binding reactivity, as indicated by geometric mean titer units on the y axis of each bar plot, was measured on the individual TCR-ß peptides listed on the x axis. The peptides were numbered by how they would correspond from the amino to the carboxyl terminus of the complete protein sequence. OR1: black bar; OR2: grey bars; OR5: white bars; Syn 2H-11: stippled bars.



View larger version (24K):
[in this window]
[in a new window]
  		Figure 2. Anti-TCR mAbs binding to TCR-{alpha} peptides. Binding reactivity, as indicated by geometric mean titer units on the y axis of each bar plot, was measured on the individual TCR-{alpha} peptides listed on the x axis. The peptides were numbered by how they would correspond from the amino to the carboxyl terminus of the complete protein sequence. Bars are as indicated in legend to Fig. 1



View larger version (24K):
[in this window]
[in a new window]
  		Figure 3. Anti-TCR mAbs binding to Mcg {lambda} peptides. Binding reactivity, as indicated by geometric mean titer units on the y axis of each bar plot, was measured on the individual Ig {lambda} Mcg peptides listed on the x axis. The peptides were numbered by how they would correspond from the amino to the carboxyl terminus of the complete protein sequence. Bars are as indicated in legend to Fig. 1 .

Geometric mean titers from these studies ranged from 5 to 150. These numbers relate the various degrees of binding reactivity the anti-TCR mAb supernatants demonstrated on the test peptides. The differences seen in the titer reactivities are related to antibody affinity for peptides. Relative binding affinities of mAb cannot be accurately compared, however, because of the different concentrations of mAb expressed in the recovered supernatants.

With respect to the binding profiles observed on the TCR and Ig {lambda} peptides, it may be likely OR2, OR5, and Syn 2H-11 are polyreactive antibodies. Similar ELISA studies were conducted with the anti-TCR mAbs on whole protein test antigens. These experiments helped determine whether IgM mAbs derived from RA patients were rheumatoid factors. The anti-TCR mAbs were examined on fetuin, BSA, ovalbumin, thyroglobulin, and pooled human polyclonal IgG. Reactivity to all test proteins by the anti-TCR mAbs was nonexistent even at the most concentrated titers. The titer of the human RA plasma control to polyclonal human IgG at an OD of 0.5 carried out to 1 in 12,000.

Anti-TCR mAbs display diverse binding properties on TCR CDR1 spectratype peptides
The anti-TCR mAbs were tested on a set of Vß homologue peptides corresponding to the Vß CDR1 sequence (CDR1 homologs). In the set of CDR1 homologs, the ß8.1 peptide is the designated nomenclature for the selecting ß3 peptide. The CDR1 peptides range from 33% (11.1) to 80% (6.5) in homology to ß3 (Table 4) . OR1, defined tentatively as a monospecific antibody, bound to 10 of the 24 CDR1 homologs. Mean titers to peptides 10.1, 16.1, and 21.1 were the highest (50–100 geometric mean units) and suggest greater binding affinity to these peptides (Fig. 4 ). Sequence examination of these peptides show that all contained a PIXXH motif; those that did not were unable to generate measurable titers. A notable exception is peptide ß6.1, which contains the PIXXH motif and is 69% homologous to ß8.1, but was not bound by OR1. Presumably this lack of binding is attributed to inhibition residues not expressed in the other peptides. Alternatively, some peptides (ß15.1 and ß19.1) lacking the PIXXH motif could still elicit a small but measurable titer (Table 4) . Although distinct patterns are apparent in the peptides that elicited a strong OR1 response, the definitive impact of other key residues not identified in these observations is considered crucial for the degree of binding.



View larger version (33K):
[in this window]
[in a new window]
  		Figure 4. Anti-TCR mAbs binding to CDR1 (ß3) homologs. Binding reactivity, as indicated by geometric mean titer units on the y axis of each bar plot, was measured on the individual CDR1 peptides listed on the x axis. Bars are as indicated in legend to Fig. 1 .

The reactivity of OR2 on the CDR1 homologs was the most extreme among the mAbs. It bound to 20 of the 24 CDR1 homologs in mean titer from 4 to > 400 units. OR2 bound to 13 of the peptides with a greater titer measurement (ß6.1, ß6.5, ß9.1, ß10.1, ß11.1, ß12.1, ß13.1, ß14.1, ß15.1, ß16.1, ß17.1, ß19.1, and ß21.1) than the selecting ß8.1 peptide (Fig. 4) . The diverse binding profile of this antibody on the CDR1 homologs may be enhanced by its epitope promiscuity. Because of this unpredictable property, it is impossible to distinguish where cross-reactivity and heteroclytic reactivity overlap. Key residues are not present in finite sequence motifs, but interchange such that specific binding interactions are not disrupted.

The mAbs OR5 and Syn 2H-11 demonstrated a range of binding reactivities similar to that of OR1. OR5 bound similarly to ß6.5, ß10.1, ß12.1, ß15.1, ß16.1, and ß23.1, with the highest mean titer for ß10.1. Syn 2H-11 bound to ß10.1, ß12.1, ß15.1, ß16.1, and ß23.1 with the highest mean titer for ß10.1, but notably high (> 50 mean titer units) for ß12.1, ß15.1, and ß16.1. Syn 2H-11 bound to ß9.1, ß14.1, ß20.1, and ß21.1 (Fig. 4) . Again, the motifs of the peptides recognized by OR5 and Syn 2H-11 share the common PIXXH sequence, implying these peptides may be more permissive to binding by the different anti-TCR mAbs despite variable sequences in the heavy and light chain antigen combining site.

Examining binding reactivity of epitope promiscuous OR2 on TCR FR3 spectratype peptides
Anti-TCR mAbs were tested on a set of peptides homologous to ß8, which corresponds to the FR3 region of the human Vß TCR. The five peptides examined ranged from 25% to 62.5% in sequence homology to ß8 (Table 5) . The results from these experiments revealed that only OR2 was capable of binding to the FR3 homologs. Examination of the titers indicated that the strongest binding interactions occurred on ß5.2 and ß17.1, which were close to 400 mean titer units. Sequence comparisons between these two peptides bear little resemblance except for the AXYXCA in the amino-terminal end corresponding to the ß8 peptide. The titer data suggest, however, that this motif has little effect on the binding interaction between OR2 and the FR3 homologs given the relatively low mean titers generated against the other peptides (Fig. 5 ).



View larger version (45K):
[in this window]
[in a new window]
  		Figure 5. OR2 anti-TCR mAbs binding to the FR3 (ß8) homologs. Binding reactivity, as indicated by geometric mean titer units on the y axis of each bar plot, was measured on the individual FR3 peptides listed on the x axis.

Unrelated peptides share the same binding site in epitope promiscuous anti-TCR mAb OR2
In studies of the property of epitope recognition promiscuity, we focused on mAb OR2 due to its diverse binding profile on TCR and Ig {lambda} peptides. As mentioned, the anti-TCR mAbs were selected on the recombinant single-chain T cell receptor (scTCR) and ß3 peptide. The following experiments examined the nature of the OR2 binding site by using selected TCR antigens to inhibit its binding in competition ELISAs. The constant antibody binding site concentration in these assays was estimated at 2.0 x 10-8 M. The first set of experiments, illustrated in Fig. 6 , shows that OR2 binds to these antigens in a specific manner. The ß3 peptide and scTCR interact with the OR2 combining site to the effect that increasing the antigen concentration can inhibit antibody binding. Figure 6b, c examines whether inhibitions with soluble antigens could be carried out against the same coating antigens. It was estimated that a 0.34 x 10-6 M concentration of scTCR and an 8.5 x 10-6 M concentration of ß3 were sufficient to block OR2 binding by 50%. In repeated experiments increasing molar amounts of the ß1 control peptide were unable to inhibit the level of binding of OR2 to immobilized ß3 (Fig. 6b ). The scTCR and the ß3 peptide could also cross-inhibit against alternated immobilized antigens. An approximate scTCR concentration of 0.6 x 10-6 M was sufficient to block OR2 binding to ß3-coated plates by 50% (Fig. 6a ) and 21.0 x 10-6 M of ß3 was sufficient to block OR2 binding to scTCR-coated plates by 50% (Fig. 6d ).



View larger version (22K):
[in this window]
[in a new window]
  		Figure 6. Inhibition of OR2 binding to immobilized ß3 peptide and scTCR in ELISA with soluble ß3 and scTCR. Increasing molar amounts of soluble antigen (x axis) were mixed with an invariable antibody dilution (2.0x10-8 M). Concentration of immobilized peptide and scTCR was 2.5 x 10-6 M and 3.0 x 10-8 M. Inhibition of antibody binding to immobilized antigen on ELISA plate was measured as a percentage (y axis). a) Soluble scTCR ({circ}) on ß3-coated ELISA; b) soluble ß3 (•) and ß1 negative control peptide ({triangleup}) on ß3-coated ELISA; c) soluble scTCR on scTCR-coated ELISA; and d) soluble ß3 on scTCR-coated ELISA.

Of the TCR-ß peptides, OR2 reacted with the highest titer to ß15, indicating that the OR2 binding site has a greater specific affinity for ß15 than it does for the selecting peptide (Fig. 1) . Similar inhibition experiments therefore were conducted with the TCR-ß15 peptide. The ß15 peptide, however, is completely unrelated to ß3 sequentially (Table 1) . On ß15-coated ELISAs, a concentration of ~0.3 x 10-6 M of scTCR was able to inhibit OR2 binding by 50%. A molar concentration of ~1.25 x 10-6 M of soluble ß15 was sufficient to accomplish the same level of OR2 inhibition on ß15-coated plates (Fig. 7 a). The ß1 specificity control peptide was unable to inhibit OR2 binding to ß15-coated plates (Fig. 7b ). Inhibition results for soluble ß15 on plates coated with ß3 and scTCR were similar. A molar concentration of 3.1 x 10-6 M of ß15 was required to attain 50% inhibition on ß3-coated plates and 1.0 x 10-6 M of ß15 could block 50% OR2 binding to scTCR-coated ELISAs (Fig. 7c, d ). In assays using soluble ß3 to inhibit OR2 binding to ß15-coated plates, no OR2 inhibition could be measured. These results clearly demonstrate that the anti-TCR mAb OR2 binds to the cognate or selecting antigens (scTCR peptide and ß3) in a specific manner and that the binding site is shared by these two antigens. OR2 binds specifically to the noncognate, nonselected antigen ß15. These data are representative of the overall results obtained from inhibition studies.



View larger version (20K):
[in this window]
[in a new window]
  		Figure 7. Inhibition of OR2 binding to immobilized ß15 peptide and scTCR in ELISA with soluble ß15 and scTCR. Increasing molar amounts of soluble antigen (x axis) were mixed with an invariable antibody dilution (2.0x10-8 M). Concentration of immobilized peptide and scTCR was 2.5 x 10-6 M and 3.0 x 10-8 M. Inhibition of antibody binding to immobilized antigen on ELISA plate was measured as a percentage (y axis). a) Soluble scTCR ({circ}) on ß15-coated ELISA; b) soluble ß15 ({diamondsuit}) and ß1 negative control peptide ({triangleup}) on ß15-coated ELISA; c) soluble ß15 on ß3-coated ELISA; and d) soluble ß15 on scTCR-coated ELISA.

Determination of the dissociation constant (KD) of OR2 on unrelated TCR peptides and scTCR
Inhibition studies carried out with OR2 give an indication of the affinity this mAb has for specific TCR epitopes, which can be determined in several ways from these experiments. If a specific molar concentration of soluble antigen can block the binding of OR2 to an antigen-coated well to the degree that the optical density units are 1/2 the absorbance of an equivalent OR2 concentration in control wells without soluble antigen, then theoretically 1/2 of the total OR2 binding sites could be considered occupied. This can be directly applied to the Scatchard equation (28) . OR2 appears to have a very weak affinity for the ß3 peptide, but has about a 20-fold greater affinity for the ß15 peptide and a 70-fold greater affinity for scTCR when determining the working KD from inhibitions conducted on scTCR (cognate antigen) -coated ELISAs (Table 6) . Although ß3 peptide is a specific epitope for OR2, it does not appear to be the cognate epitope based on relative affinities.


View this table:
[in this window]
[in a new window]
  		Table 6. Estimated molar concentrations of soluble antigen required to inhibit OR2 binding immobilized antigen by 50%

Inhibition ELISAs have been used to determine affinity constants of antibodies in earlier studies. Results are comparable to those obtained from fluorescence transfer and immunoprecipitation experiments. The values of free vs. bound antibody sites were translated from comparing absorbance measurements of antibody with soluble antigen to those of antibody without inhibiting antigen (29) . This system could be applied to the inhibition studies performed. Scatchard plots were generated from inhibition experiments conducted with soluble ß3, ß15, and scTCR on ß3-coated ELISAs (Fig. 8 ). Relative dissociation constants for OR2 were comparable to the working KD values obtained for 50% binding in that the plots demonstrate the lowest antibody affinity for ß3, followed by ß15, and the highest affinity for scTCR.



View larger version (12K):
[in this window]
[in a new window]
  		Figure 8. Scatchard plots of the binding of inhibiting soluble antigens ß3, ß15, and scTCR to OR2 on ß3-coated ELISA. Data from the inhibition experiments was used to calculate KD values for OR2 binding to ß3 (•), ß15 ({diamondsuit}), and scTCR ({circ}) (29) .


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
These studies confirm that epitope recognition promiscuity is a property of some natural antibodies and can manifest in various degrees in monoclonal natural antibodies. Epitope promiscuity can be distinguished by several features: 1) reactivity to one or more epitopes unrelated to the cognate antigen, 2) nonreactive to proteins that often represent antigens bound by polyreactive antibodies, and 3) antigens compete for the same binding site, but affinities can vary. The four mAbs examined in these studies were selected for on the same TCR antigens, but all possess unique V region sequences, particularly in the CDR3 of the heavy chain (27) . Reactivity against the overlapping TCR-ß, -{alpha}, and Mcg {lambda} peptides was displayed in unique patterns ranging from monospecific binding (OR1), to restricted epitope promiscuous binding (OR5 and Syn 2H-11), to highly epitope promiscuous binding (OR2).

We have examined the differences between OR1 and OR2, two mAbs selected on the same antigens but exhibiting contrasting epitope binding properties. OR1 reacted in a classical sense to the ‘cognate’ antigen used in the selection procedure by binding in a restricted fashion. OR1 also bound to the {alpha}3 and {alpha}4 peptides but not to the homologous region of the Mcg {lambda} chain. The property of epitope promiscuity must be challenged in this scenario due to the existence of shared sequence between the ß3 (WYRQT) and the {alpha}3 and {alpha}4 peptides (WYVQY). This conjecture is supported in the molecular models of the complete TCR-{alpha} and TCR-ß chain Fig. 9 . The regions highlighted in yellow represent the peptide sequences to which OR1 reacted. The structural motif of the V{alpha}/ß TCR to which OR1 binds spans from the first amino-terminal ß-strand through the exposed loop representing the CDR1 back through the second ß-strand and part of the next {alpha} helix turn representing part of the FR2. It is more likely, based on the sequence homologies and the models, that the binding interaction of OR1 to the {alpha} peptides is a property of cross-reactivity. OR1 possesses a sequence motif (RFLEW) in its heavy chain CDR3 (27) that has been described in 4 of 50 known polyreactive antibodies and none of the 2500 human antibodies believed to be monoreactive (30) . These findings suggest that OR1 may be an exception to the current research and, moreover, represents a caveat in assessing binding reactivity from sequence motifs alone.



View larger version (70K):
[in this window]
[in a new window]
  		Figure 9. Molecular models of the complete TCR {alpha} and ß chains with highlighted regions of OR1 binding. Segments on the three-dimensional structures corresponding to 16-mer peptides bound to OR1 antibody. Atomic coordinates are taken from the Protein Data Bank (PDB) for examples of a TCR {alpha} chain (PDB#1ao7) and a TCR ß chain (PDB#1 bec). Ribbon models are drawn with the program MOLMOL (32) . Secondary structures are represented by directional arrows (ß-strands) or spirals (helices). Putative epitopes represented by peptides bound by OR1 are highlighted in yellow. They correspond to complete CDR1 and amino-terminal parts of the contiguous FR2 segments.

OR2 represents the most extreme example of epitope recognition promiscuity in these studies. This mAb reacted to unrelated TCR peptides, often with what appeared to be comparable affinities (based on direct binding titer data). Figure 10 shows the same TCR-{alpha} and -ß model and a model for the complete Mcg {lambda} chain with highlighted regions of OR2 binding. The structure of Mcg was determined crystallographically (23) and has been used as a template to model the structure of the YT35 ß chain (31) . OR2 reacted to two conspicuous regions in the variable portions of the structures highlighted in yellow and cyan. These regions correspond to the CDR1 and part of the FR2 (yellow) and the carboxyl-terminal end of the FR3 to the CDR3 (cyan). OR2 reacted to the J region (FR4) of the Mcg {lambda} chain highlighted in royal blue. The major constant region structures to which OR2 bound are highlighted in violet, green, and orange. OR2 bound to the various unrelated secondary structures in the constant regions of the models. This observation is supported by the lack of recognizable homology between the TCR-{alpha}, TCR-ß, and Ig {lambda} peptide sequences corresponding to the different constant regions. In summary, OR2 binds to different regions of the constant and variable TCR and Mcg light chain. Making the distinction between cross-reactivity and epitope promiscuity for binding to V region segments is difficult, but becomes obvious when examining binding to constant regions.



View larger version (38K):
[in this window]
[in a new window]
  		Figure 10. Molecular models of the complete TCR {alpha} and ß chains, and the complete Mcg {lambda} chain with highlighted regions of OR2 binding. Segments on the three-dimensional structures corresponding to 16-mer peptides bound to OR2 antibody. Atomic coordinates are taken from the PDB for examples of a TCR {alpha} chain, a TCR ß chain and the two human Ig {lambda} chains (Mcg, PDB#2 Mcg). Putative epitopes represented by peptides bound by OR2 are highlighted in yellow, cyan, royal blue, violet, green, and dark orange. They correspond to the complete CDR1 and amino-terminal and parts of the contiguous FR2 region, the carboxyl-terminal end of the FR3 to the CDR3, the J region (FR4) of the Mcg {lambda} chain, and to constant region portions of the TCR {alpha} and ß and Mcg {lambda} chains.

Experiments investigating binding of the anti-TCR mAbs on the CDR1 (ß3) and FR3 (ß8) homologs expand on epitope mapping studies. These data provide two important details about the anti-TCR mAbs used in these experiments and the nature of epitope recognition promiscuity. First, they support the earlier findings that binding to human CD3+ PBMCs occurs only on restricted Vß subsets (27) (binding to V{alpha} subsets is unknown). Even the epitope promiscuous mAb OR2 did not bind to every CDR1 homologue (Fig. 4) . Second, the degree of epitope promiscuity may have a profound effect on the level of cross-reactivity an antibody can demonstrate against homologous epitopes. It is conceivable that these properties can synergize to enhance the binding reactivity such as what was observed with OR2.

The mAb OR2 best demonstrates the property of epitope recognition promiscuity. In competition ELISA studies, OR2 bound to the epitopes on which it was selected (ß3 and scTCR) as well as an unrelated epitope (ß15) in a specific manner. From the analysis of these inhibition studies, it is hypothesized that all three of the antigens compete for the same binding site in OR2. The inhibition studies reveal that OR2 appears to have a very low affinity for the selecting ß3 peptide. Considerably larger molar concentrations of ß3 were required to block OR2 binding to coated ELISAs. These findings suggest that the binding to peptide ß3 used in the selection represents a cross-reaction and that the major specificity of OR2 is directed toward a different epitope. In contrast to the inhibitions using ß3, the three-domain constant and variable scTCR could inhibit OR2 binding to antigen-coated wells at far smaller molar concentrations than the ß3 and ß15 peptides (Figs. 6 , 7 ; Table 6 ). The sequence for ß3 is contained in the scTCR sequence, but it appears that OR2 reacts against another epitope(s) with greater affinity. Most likely it shows preference for conformational determinants of the intact construct. OR2 is also inhibited at lower concentrations of ß15 than ß3. The two peptides, however, are unrelated sequentially and biochemically. The estimated pI for ß3 is 9.96 and the estimated pI for ß15 is 5.18. The ß3 peptide contains 31.25% nonpolar residues, 50% polar residues, and 18.75% basic residues, whereas the ß15 peptide contains 37.5% nonpolar residues, 25% polar residues, 18.75% acidic residues, and 18.75% basic residues (Table 1) .

We estimated the KD for OR2 on the ß3 peptide, ß15 peptide, and the scTCR using data from the inhibition studies. This method has been used with monoclonal IgG (29) and IgM molecules (33) . Results from this study appear to be representative of what was expected. OR2 demonstrated a very low binding affinity KD for its selecting antigen, the ß3 peptide (~ 2.5x10-5 M). The affinity of OR2 for ß15 (~ 4.0x10-6 M) was significantly higher, but still considered a relatively low antibody binding dissociation constant. OR2 demonstrated an affinity for the scTCR (~ 3.0x10-8 M) that is considered to be relatively high for antibody binding. IgM molecules are generally considered low-affinity Igs when their binding sites are compared with that of induced IgG, but these molecules compensate low affinity with high avidity due to a higher number of binding sites. This compensation establishes a protective role for IgM in the immune system (34 35 36) and has been demonstrated in IL-2 inhibition experiments on antigen activated T cells with OR2, OR5, and Syn 2H-11 (37) .

The complex property of epitope recognition promiscuity introduces a new perspective on epitope binding interactions for Igs and TCRs; it is apparent, therefore, that new algorithms are required to successfully predict the binding relationship that would occur between an idiotope and its epitope(s). Epitope recognition promiscuity is a property of antibodies found in all species of vertebrates, from sharks to humans, and is recognized as a conserved feature of the combinatorial immune system including TCRs and Igs (8) . The variation in Ig gene products ranging from monospecific to polyreactive is not considered a random recombinatorial phenomenon. Although antibodies from these studies may possess low affinity for their respective epitopes, the binding interactions they carry out are still specific. Low-affinity antibodies are important because they can discriminate among antigens better than high-affinity antibodies (1) . Moreover, binding specificities for different epitopes by antibodies with different V region sequences on the same molecular structures represent further attestation to the overlapping functionality of the immune system.


   ACKNOWLEDGMENTS
 
This work was supported in part by contract 5–038 to J.J.M. from the Arizona Disease Control Research Commission and by grants OCAST OARS ARO1.2-017 and OCAST HR00-093 to A.B.E. We thank Daniel Smith for preparing the color figures.

Received for publication November 12, 2001. Revision received January 28, 2002.
   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Van Regenmortel, M. H. V. (1998) From absolute to exquisite specificity. Reflections on the fuzzy nature of species, specificity and antigenic sites. J. Immunol. Methods 216,37-48[CrossRef][Medline]
  2. Kazatchkine, M. D., Kaveri, S. V. (2001) Immunomodulation of autoimmune and inflammatory diseases with normal polyspecific human IgG. N. Engl. J. Med. 345,747-755[Free Full Text]
  3. Burnet, F. M. (1959) The Clonal Selection Theory of Aquired Immunity Cambridge University Press Cambridge, England.
  4. Berzofsky, J. A., Schechter, A. N. (1981) The concepts of crossreactivity and specificity in immunology. Mol. Immunol. 19,751-763
  5. Casali, P., Notkins, A. L. (1989) Probing the human B-cell repertoire with EBV: polyreactive antibodies and CD5+ B lymphocytes. Annu. Rev. Immunol. 7,513-535[Medline]
  6. Casali, P., Prabhakar, B. S., Notkins, A. L. (1988) Characterization of multi-reactive autoantibodies and identification of leu-1+ B lymphocytes as cells making antibodies binding multiple self and exogenous molecules. Int. Rev. Immunol. 3,17-45[Medline]
  7. Marchalonis, J. J., Robey, I., Schluter, S. F., Yocum, D. E. (2000) Epitope promiscuity of human monoclonal antibodies to human T cell receptor combining site determinants. Appl. Biochem. Biotechnol. 83,31-52[CrossRef][Medline]
  8. Marchalonis, J. J., Adelman, M. K., Robey, I. F., Schluter, S. F., Edmundson, A. B. (2001) Exquisite specificity and peptide epitope recognition promiscuity, properties shared by antibodies from sharks to humans. J. Mol. Rec. 14,110-121
  9. Marchalonis, J. J., Garza, A., Landsperger, W. J., Schluter, S. F., Wang, A. C. (1997) Binding of human IgG myeloma proteins to autologous T-cell receptor determinants. Crit. Rev. Immunol. 17,497-506[Medline]
  10. Keitel, T., Kramer, A., Wessner, H., Scholz, C., Schneider-Mergener, J., Hohne, W. (1997) Crystallographic analysis of anti-p24 (HIV-1) monoclonal antibody cross-reactivity and polyspecificity. Cell 91,811-820[CrossRef][Medline]
  11. Kramer, A., Keitel, T., Winkler, K., Stocklein, W., Hohne, W., Schneider-Mergener, J. (1997) Molecular basis for the binding promiscuity of an anti-p24 (HIV-1) monoclonal antibody. Cell 91,799-809[CrossRef][Medline]
  12. Marchalonis, J. J., Schluter, S. F., Bernstein, R. M., Hohman, V. S. (1998) Antibodies of sharks: revolution and evolution. Immunol. Rev. 166,103-122[CrossRef][Medline]
  13. Marchalonis, J. J., Adelman, M. K., Zeitler, B. J., Sarzin, P. M., Jaqua, P. M., Schluter, S. F. (2001) Evolutionary factors in the emergence of the combinatorial antibody repertoire. Adv. Exp. Med. Biol. 484,13-30[Medline]
  14. Seidl, K. J., Wilshire, J. A., MacKenzie, J. D., Kantor, A. B., Herzenberg, L. A., Herzenberg, L. A. (1999) Predominant VH genes expressed in innate antibodies are associated with distinctive antigen-binding sites. Proc. Natl. Acad. Sci. USA 96,2262-2267[Abstract/Free Full Text]
  15. Adib-Conquy, M., Gilbert, M., Avrameas, S. (1998) Effect of amino acid substitutions in the heavy chain CDR3 of an autoantibody on its reactivity. Int. Immunol. 10,341-346[Abstract/Free Full Text]
  16. Weller, S., Conde, C., Knapp, A. M., Levallois, H., Gilfillan, S., Pasquali, J. L., Martin, T. (1997) Autoantibodies in mice lacking terminal deoxynucleotidyl transferase: evidence for a role of N region addition in the polyreactivity and in the affinities of anti-DNA antibodies. J. Immunol. 159,3890-3898[Abstract]
  17. Dehghanpisheh, K., Bona, C. A., Marchalonis, J. J. (1996) Peptide epitope binding specificity and V{kappa} and VH gene usage in a monoclonal IgM natural autoantibody to T cell receptor CDR1 from a viable motheaten mouse. Immunol. Invest. 25,241-252[Medline]
  18. Ichiyoshi, Y., Casali, P. (1994) Analysis of the structural correlates for antibody polyreactivity by multiple reassortments of chimeric human immunoglobulin heavy and light chain V segments. J. Exp. Med. 180,885-895[Abstract/Free Full Text]
  19. Adib-Conquy, M., Gilbert, M., Christodoulou, C., Avrameas, S. (1994) Reactivity and structure of a mouse anti-F(ab')2 IgM. Comparison of its variable region sequences with those of a structurally close polyreactive natural IgM. Mol. Immunol. 31,555-562[CrossRef][Medline]
  20. Chen, C., Stenzel-Poore, M. P., Rittenberg, M. B. (1991) Natural auto- and poly-reactive antibodies differing from antigen-induced antibodies in the H chain CDR3. J. Immunol. 147,2359-2367[Abstract]
  21. Yanagi, T., Yoshikai, Y., Leggett, K., Clark, S., Aleksander, I., Mak, T. W. (1984) A human T cell specific cDNA clone encodes a protein having extensive homology to immunoglobulin chains. Nature (London) 308,145-149[CrossRef][Medline]
  22. Yanagi, Y., Chan, A., Chin, B., Minden, M., Mak, T. W. (1985) Analysis of cDNA clones specific for human T cells and the alpha and beta chains of the T-cell receptor heterodimer from a human T-cell line. Proc. Natl. Acad. Sci. USA 82,3430-3434[Abstract/Free Full Text]
  23. Edmundson, A. B., Ely, K. R., Abola, E. E., Schiffer, M., Pangiotopoulos, N. (1975) Rotational allomerism and divergent evolution of domains in immunoglobulin light chains. Biochemistry 14,3953-3962[CrossRef]
  24. Arden, B., Clark, S. P., Kabelitz, D., Mak, T. W. (1995) Human T-cell receptor variable gene segment families. Immunogenetics 42,455-500[Medline]
  25. Marchalonis, J. J., Kaymaz, H., Dedeoglu, F., Schluter, S. F., Yocum, D. E., Edmundson, A. B. (1992) Human autoantibodies reactive with synthetic autoantigens from T-cell receptor ß chain. Proc. Natl. Acad. Sci. USA 89,3325-3329[Abstract/Free Full Text]
  26. Lake, D. F., Helgerson, S., Landsperger, W. J., Marchalonis, J. J. (1997) Physical and epitope analysis of a recombinant human T-cell receptor V alpha/V beta construct support the similarity to immunoglobulin. J. Prot. Chem. 16,309-320[CrossRef][Medline]
  27. Robey, I. F., Schluter, S. F., Yocum, D. E., Marchalonis, J. J. (2000) Production and characterization of monoclonal IgM autoantibodies specific for the T-cell receptor. J. Prot. Chem. 19,9-21[CrossRef][Medline]
  28. Scatchard, G. (1949) The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51,660-672[CrossRef]
  29. Friguet, B., Chaffotte, A. F., Djavadi-Ohaniance, L., Goldger, M. E. (1985) Measurements of the true affinity constant in solution of antigen-antibody complexes by enzyme-linked immunosorbent assay. J. Immunol. Methods 77,305-319[CrossRef][Medline]
  30. Deng, Y.-J., Notkins, A. L. (2000) Molecular determinants of polyreactive antibody binding: HCDR3 and cyclic peptides. Clin. Exp. Immunol. 119,69-76[CrossRef][Medline]
  31. Kaymaz, H., Dedeoglu, F., Schluter, S. F., Edmundson, A. B., Marchalonis, J. J. (1993) Reactions of anti-immunoglobulin sera with synthetic T cell receptor peptides: implications for the three-dimensional structure and function of the TCR beta chain. Int. Immunol. 5,491-502[Abstract/Free Full Text]
  32. Koradi, R., Billeter, M., Wuthrich, K. (1996) MOLMOL: a program display and analysis of macromolecular structures. J. Mol. Graphics. 14,51-55[CrossRef][Medline]
  33. Diaw, L., Magnac, C., Pritsch, O., Buckle, M., Alzari, P. M., Dighiero, G. (1997) Structural and affinity studies of IgM polyreactive natural autoantibodies. J. Immunol. 158,968-976[Abstract]
  34. Ochsenbein, A. F., Fehr, T., Lutz, C., Suter, M., Brombacher, F., Hengartner, H., Zinkernagel, R. M. (1999) Control of early viral and bacterial distribution and disease by natural antibodies. Science 286,2156-2159[Abstract/Free Full Text]
  35. Boes, M. (2000) Role of natural and immune IgM antibodies in immune responses. Mol. Immunol. 37,1141-1149[CrossRef][Medline]
  36. Hurez, V., Kazatchkine, M. D., Vassilev, T., Ramanathan, S., Pashov, A., Basuyaux, B., de Kozak, Y., Bellon, B., Kaveri, S. V. (1997) Pooled normal human polyspecific IgM contains neutralizing anti-idiotypes to IgG autoantibodies of autoimmune patients and protects from experimental autoimmune disease. Blood 90,4004-4013[Abstract/Free Full Text]
  37. Robey, I. F., Akporiaye, E., Schluter, S. F., Yocum, D. E., Marchalonis, J. J. (2001) Human monoclonal natural autoantibodies against the T cell receptor inhibit IL-2 production in murine T cells. Immunology In press




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
Services
Right arrow Email this article to a friend
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
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by ROBEY, I. F.
Right arrow Articles by MARCHALONIS, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by ROBEY, I. F.
Right arrow Articles by MARCHALONIS, J. J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS