Antibodies

Adverse immunological reactions to self and foreign antigens and other situations of abnormal lymphocyte growth are the cause of autoimmune disease, hypersensitivity and asthma, lymphoid cancer and transplant rejection.

We now know that the decisions leading to lymphocyte survival or death are decided not just by antigen-binding but also by activating or inhibitory coreceptors that provide positive or negative selection signals and tune cells to vital cues in their environment. Understanding these processes has led to the development of inhibitory antibodies or fusion proteins that mask the ligands of the activating receptors. The parallel strategy of developing agonistic antibodies that directly activate inhibitory pathways in lymphocytes has not been attempted, despite a theoretical advantage that these agents might be more potent and easily targeted to lymphocyte subsets and inflammatory locations. The creation of a new class of antibody superagonists capable of activating inhibitory receptors would, we think, be a major therapeutic advance.

Insights into the effects of supergonists arrived somewhat serendipitously. We were sponsored by industry to determine the crystal structure of CD28, but could not deglycosylate it. We wondered if suspending the protein in a crystal lattice with Fab fragments of antibodies would allow us to leave the sugars on the protein, and this eventually turned out to be a good idea. We approached Prof Thomas Hunig of the University of Wurzburg who we knew worked with CD28 antibodies. Unexpectedly, Prof Hunig was able to supply a superagonistic antibody, 5.11A1, which we thought would significantly ratchet up interest in the structure of an antibody complex if we could obtain it making the structural study even more exciting. We subsequently learnt from Prof Hunig’s work that a simple relationship exists between epitope location and mitogenic potential for these antibodies. What our crystal structure of the CD28/5.11A1 Fab complex eventually showed (1) (Fig. 1) is that superagonists bind at the “side” of the homodimer, whereas non-superagonists bind at the top. This led us to suggest that the greater potency of superagonists arises simply from their ability to more effectively exclude large phosphatases, e.g. CD45 or CD148, from the vicinity of CD28, favouring its phosphorylation (2) (Fig. 2).

An important corollary of these ideas is that any immobilized antibody binding to a membrane-proximal epitope on a receptor with tyrosine phosphorylation motifs should induce signalling by that receptor. We are now testing this idea by preparing antibodies against the human inhibitory receptor, PD-1, which is a small signalling protein with so-called ITIM and ITSM signalling motifs and is among the most potent inhibitory proteins expressed on T cells. Interest in this molecule increased dramatically following the discovery that it is responsible for the ‘exhausted’ phenotype acquired by T cells in the course of chronic viral infections. This suggested the possibility that antibody-induced superagonistic PD-1 signalling could be used clinically to impose an ‘exhausted’ phenotype on auto-reactive T cells. If the work with PD-1 is promising we will extend it to other targets.

For an animation of the antibody signalling idea, please see Movie 1.

Fig. 1: Positions of the CD28 homodimer bound by superagonistic (5.11A1) and conventional (7.3B6) antibodies

The superagonists (red) bind at the “side” of the homodimer (yellow), whereas the non-superagonists (green) bind at the top.

Fig. 1

Fig. 2: Broad implications of the KS model

(a) Differences in the dimensions of the complexes formed by CD28 and conventional and superagonistic monoclonal antibody (mAb) to CD28, demonstrated by crystallographic and cryoelectron microscopy analyses. (b) It is proposed that in vivo, the more-compact complexes formed by superagonists more effectively exclude large proteins such as phosphatases, leading to more-potent signaling by CD28 (left). The apparent sensitivity of CD28 triggering to size-dependent effects suggests that it occurs in a KS-like way (right).

Fig. 2
Movie 1

Citations:

  1. Evans EJ, Esnouf RM, Manso-Sancho R, Gilbert RJ, James JR, Yu C, Fennelly JA, Vowles C, Hanke T, Walse B, Hünig T, Sørensen P, Stuart DI, Davis SJ. (2005) Crystal structure of a soluble CD28-Fab complex. Nat Immunol. 6, 271-9.
  2. Davis SJ, van der Merwe PA. (2006) The kinetic-segregation model: TCR triggering and beyond. Nat Immunol. 7, 803-9.

Antibody Papers