Antibody Recognition of Antigen
David Marcey
© 2007

I. Introduction
II. Recognition of a globular antigen
III. Recognition of a peptide antigen
IV. Recognition of a hapten antigen
V. References


This exhibit displays molecules in the left part of the screen, and text that addresses structure-function relationships of the molecules in the right part (below). Use the scrollbar to the right to scroll through the text of this exhibit.

To evoke renderings of the molecule that illustrate particular points, click the radio buttons:

To reset the molecule, use the load/reset buttons:

If you resize your browser window, simply refresh the page in order to restore proper viewing.

I. Introduction

Note: this tutorial assumes knowledge of the basics of immunoglobulin structure - see An Introduction to Immunoglobulin Structure.

The staggeringly large repertoire of antibodies with different antigen-binding specificity is the basis for the immune system's ability to recognize virtually all foreign antigens. The structural basis for the repertoire can be found in the variations of VL and VH structure in different antibodies. These two amino-terminal domains lie at each tip of the branches of the immunoglobulin. The VL and VH hypervariable regions (CDR's) project from the end of their domains, ready for antigen binding. This tutorial will use three examples of antibody-antigen complexes to illustrate some structural features of antibody recognition of epitope. First we'll look at antibody bound to a globular protein antigen. Then we'll examine the mechanisms of peptide antigen binding. We'll finish by studying antibody recognition of a small antigenic molecule, a hapten.

return to beginning of the exhibit

II. Recognition of a globular antigen


To the left is a crystal structure (at 2.5 Å resolution) of a Fab fragment (VLCL-VHCH1) from an anti-lysozyme antibody complexed with a globular antigen, hen egg white lysozyme (HEL) (Fischmann, et al., 1991).

The variable VL and VH domains make extensive contacts with HEL across a broad, flat surface at the Fab tip.

Viewing just the VL and VH domains plus HEL, we can examine these contacts more closely.

Potential hydrogen bonding between donor-acceptor pairs on 5 VL residues (Tyr32, Tyr50, Thr53, Phe91, Ser93) and 3 HEL residues (Asp18, Asn19, Gln121) has been identified .

The VH domain has 5 residues (Gly53, Asp54, Asp100, Tyr101, Arg102) that can form H-bonds with 7 residues in HEL (Gly22, Ser24, Asn27, Gly117, Asp119, Val120, Gln121) .

Note that both antibody main chain and side chain atoms contact both main and side chain atoms of the HEL epitope. Water molecules at the antibody-antigen interface (not shown) also contribute to the complex hydrogen bonding pattern between the molecules.

There is exquisite complementarity of the surfaces of the molecules at their interface. In addition to the hydrogen bonding patterns just discussed, numerous van der Waals interactions result .

The antigenic epitope is noncontiguous, comprising twelve HEL residues divided into two stretches (residues 18, 19, 22, 24, 27 and residues 117, 118, 119, 120, 121, 124, 125).

Residues in all six CDR's (VL CDR1, CDR2, CDR3 and VH CDR1, CDR2, CDR3) of the antibody participate in recognition of the HEL epitope.

The reader is encouraged to repeat this section of the tutorial, rotating the molecules in the left frame by clicking and dragging and changing the magnification using either of the buttons below. The molecules (VL and VH plus HEL) can be reset to their starting magnification and orientation.

zoom in           zoom out                    

return to beginning of the exhibit

III. Recognition of a peptide antigen


*This composite PDB file was produced by combining 1hil and 1ifh.

Rini et al. (1992) have compared the structures of a Fab fragment of a monoclonal antibody to influenza virus hemagglutinin (HA) in both ligand-bound and unliganded forms . If you are interested in hemagglutinin structure, see Viral Antigens: Influenza Hemagglutinin.

At far left is displayed the 3-D structure (at 2.8 Å resolution) of the VL and VH portions of the Fab complexed with the heptapeptide epitope from HA (liganded). To the right of this is the structure of VL and VH from the same Fab without bound antigen (unliganded). Examination of the antigen binding pocket in both structures reveals that a pronounced conformational change has occurred upon antigen binding. The pocket is deformed by antigen, closing around it.

The binding pocket deformation is mostly caused by a shift in the orientation of CDR3 of VH. Note In particular the orientation of Asp99 and Asn100 in both structures.

This structural comparison provides strong evidence in support of an induced fit model for the mechanism of antibody-antigen recognition. Induced fit may explain why some antibodies can recognize both intact protein antigens and small, free peptide epitopes from such proteins, even though an epitope embedded in a globular protein is presented in a considerably different environment from that of a free peptide. Also, an induced fit mechanism may confound some attempts to define the shape of antibody combining sites based on the structure of unliganded Fabs.

return to beginning of the exhibit

IV. Recognition of a hapten antigen

To the left is shown a 2.9 Å resolution structure of the VL and VH portions of an anti-dinitrophenyl monoclonal antibody Fab bound to its hapten antigen, DNP (Bruenger, et al., 1991).

In contrast to the binding of protein and peptide antigens, antibody binding to smaller hapten molecules usually involves the insertion of the hapten into a deep crevice or pocket between the CDR loops of the antibody. This type of binding, reminiscent of the interaction of enzyme with substrate, holds promise for the production of catalytic antibodies, using haptens that resemble a transition state of a particular substrate as antigens.

In the case at hand, the DNP can be observed to fit into the molecular canyon between VL and VH. This crevice is lined with numerous tyrosine residues .

Two tryptophan residues, one each from VL and VH, sandwich the DNP . Interestingly, neither of these trp residues accounts for the specificity of DNP binding because the germline variable region genes from which the heavy and light chains of this antibody were likely to have been derived have been characterized: both encode trp at the same positions as the trps under consideration.

However, a residue that is different from that predicted from the germline gene is tyrosine 31 from VL. The codon for this residue, presumably generated through gene rearrangement during B cell differentiation, produces a tyr that is stacked perpendicularly to the VL trp involved in sandwiching DNP. This tyr residue of the monoclonal antibody may therefore be instrumental in forming the DNP-specific binding site.

return to beginning of the exhibit

V. References

Brunger, A.T., Leahy, D.J., Hynes, T.R., Fox, R.O. (1991). 2.9 A resolution structure of an anti-dinitrophenyl-spin-label monoclonal antibody Fab fragment with bound hapten. J.Mol.Biol. 221: 239-256.

Fischmann, T.O., Bentley, G.A., Bhat, T.N., Boulot, G., Mariuzza, R.A., Phillips, S.E.V., Tello, D., and R.J. Poljak (1991). Crystallographic Refinement of the Three-dimensional Structure of the FabD1.3-Lysozyme Complex at 2.5-Å Resolution. J. Biol. Chem. 266: 12915-12920.

Padlan, E. (1994) Anatomy of the Antibody Molecule. Molecular Immunology 31: 169.

Rini, J.M., Schulze-Gahmen, U., Wilson, I.A. (1992). Structural evidence for induced fit as a mechanism for antibody-antigen recognition. Science 255: 959-965.

Schulze-Gahmen, U., Rini, J.M., Wilson, I.A. (1993). Detailed analysis of the free and bound conformations of an antibody. X-ray structures of Fab 17/9 and three different Fab-peptide complexes. J.Mol.Biol. 234: 1098-1118.

return to beginning of the exhibit