The immune response depends upon coordinated recognition both of intact and fragmented forms of protein antigens. T lymphocytes use T cell antigen receptors (TCRs) to recognize peptides associated with major histocompatibility complex (MHC) molecules. The TCR and peptide-MHC complexes meet exclusively in an immunological synapse between the T cell and peptide-MHC–presenting cell (T synapse)1. In a report in Nature, Neuberger and colleagues focus our attention on a different form of antigen presentation, one which involves intact antigens on the surface of an antigen-presenting cell (APC) interacting with the B cell antigen receptor (BCR, also known as surface immunoglobulin)2. Presentation of intact antigen as a transmembrane protein triggers formation of a BCR-dependent immunological synapse (B synapse). The B cell synapse extracts intact antigen from the APC that is then processed to generate peptide-MHC complexes for T synapse formation. The report by Neuberger's lab emphasizes the multistage “relay race”—which involves at least two synapses that are specialized for recognition of intact and degraded antigens—that is mounted by the immune system in response to soluble toxins and viruses. The description of the B synapse calls our attention to emerging work on the unique aspects of molecular interactions in synapses, integration of innate and adaptive immunity, cellular trafficking of intact antigens and viral particles and the mechanism of B cell affinity maturation.

The BCR- and TCR-mediated immunological synapses are strikingly similar in terms of antigen receptor recruitment, exclusion of negative regulators such as CD45 and the stimulation of actin polymerization (Fig. 1). A key element of the T synapse is the specific relationship between the ring of integrin adhesion molecules and the central cluster of TCRs3. (The location of adhesion molecules in the B cell synapse was not explored by Neuberger and colleagues, however, and remains to be determined.)

Figure 1: Passing the antigenic “baton” in the immunological relay race.
figure 1

(a) Antigens or immune complexes captured by DCs are presented to B lymphocytes (B synapse), which extract antigen for processing and presentation to helper T lymphocytes (T synapse). (b) In the B synapse, antigen and the BCR are clustered along with many effectors of signal transduction. Molecules that may negatively regulate BCR signal transduction are excluded from the synapsis area. (LN, lymph node; LFA-1, lymphocyte function–associated antigen 1.)

Full size image

The study2 builds on decades of investigation into the effects of antigen-binding affinity on the humoral response. It was quickly appreciated that the association of multiple antibody-binding sites in a single particle (either by natural association or artificial linkage via covalent backbones such as high molecular weight dextrans) greatly enhances the antibody response both in vitro and in vivo. This type of multipoint binding of antigens to the BCR elicits microscopic clustering and capping through signaling and actin-myosin contractility4. More recently, it has been appreciated that tethering antigens to membranes also greatly increases their potency in inducing B cell activation in vivo5. Unlike the precise geometric organization of viral particles or the regular spacing of dextrans, the organization of proteins on cell membranes is less well understood. It is clear, however, that many membrane proteins are freely mobile and thus present an ever shifting array of independent binding sites. Nonetheless, it has been directly shown that very low affinity interactions can generate organized areas of synapsis in which interactions with low solution affinities are highly efficacious6. Therefore, recognition of intact antigens in a synapse allows more efficient recognition based on very low affinity interactions, but does not allow discrimination of interactions with affinities greater than 106 M−1. How is this low-affinity ceiling reconciled with the selection of high affinity antibodies? In earlier work, Neuberger proposed that the physical nature of the antigen extraction process may itself provide discrimination of higher affinity interactions, as the resistance of a chemical interaction to applied force is related to its affinity7. Clathrin and dynamin, proteins involved in the formation of endocytic vesicles, apply significant forces to membrane receptors. This force may effectively test the affinity (in the range 106–1010 M−1) of the antibody-antigen interaction in the B synapse and provide the necessary feedback for affinity maturation.

What role does the APC play in the B synapse in vivo? Neuberger and colleagues have shown that immune complexes captured on FcR+ cells can be presented to form a B synapse2. They did this by generating a chimeric protein composed of hen egg lysozyme, green fluorescent protein and the transmembrane and cytoplasmic domains of H-2Kd, which they expressed in a tumor cell, thereby generating a model APC with which to test the concept of generically membrane-anchored antigens participating in synapse formation. It is thought that, in vivo, through Fc and complement receptors, follicular dendritic cells (FDCs) capture the immune complexes on their surface and present them to germinal center B cells. Beadlike structures called iccosomes—which contain antibody, antigen and complement products—can be identified on the dendritic processes of FDCs. Centrocytes in the germinal centers capture these complexes from the FDCs and present peptides degraded from them in order to elicit T cell help for affinity maturation8. The presentation of intact antigens as immune complexes on FDCs is thought to be important for the affinity maturation of antibodies. How-ever, early in a primary response, secreted specific antibody may not be available. In the absence of antigen-antibody complexes, other mechanisms may serve to decorate an APC with intact native antigen.

Other recent evidence points to a role for dendritic cells (DCs) in antigen presentation to primary B cells. It has been reported that DCs pulsed with soluble antigen can prime naïve B cells in T cell–dependent responses in vivo and in vitro9. Direct contact between the DCs and the B cells was required for subsequent B cell antigen presentation to T cells. DCs that were antigen-pulsed and then lysed were ineffective, which suggests that DCs play an active role in antigen priming. In these studies it was found that a fraction of the native antigen internalized by DCs was protected from intracellular degradation for at least 48 h9. In addition, a subset of DCs that traffics to primary lymphoid follicles is effective at promoting primary B cell differentiation10. DCs are also implicated in carrying intact virus to the lymphoid organs for generation of an effective humoral immune response11.

According to Neuberger's data2 antigen-specific B cells can extract membrane-bound antigens from cells such as transfected fibroblasts, which clearly are not professional antigen presenters. The polymeric nature of the membrane-displayed antigen and the close cell-cell contact provided in this system2 may permit amateur APCs to function under these circumstances. Not all antigens insert themselves into a cell membrane, however.

How might the immune system take advantage of the B cell synapse to optimize the effectiveness of encounters between B cells and soluble toxin or circulating viruses? The answer may lie, in part, in the innate mechanisms of antigen binding by professional APCs. DCs and macrophages express a variety of lectins and scavenger receptors that facilitate direct or indirect recognition of carbohydrate patterns, bacterial structures, apoptotic cells and debris associated with cell damage12. These receptors may represent a branch of the innate immune system that is specialized to capture specific classes of intact antigens, which are evolutionarily connected to pathogens. Maturing DCs have a remarkable capacity to internalize and protect intact antigens in lysosome-like compartments. With maturation of the DC, many of these antigens are degraded, but many may also be moved to the surface as intact antigens.

A number of pathogens can initiate the mannan-binding lectin and alternative pathways of complement activation. The classical and alternative pathways of complement activation all tag antigens with C3 cleavage products, which are ligands for the CD21 and CD35 complement receptors. Thus, antigens can be bound to complement receptors even in the absence of complement-fixing antibody. In addition, even antigen-naïve animals often express “natural antibodies”. These are immunoglobulin M (IgM) and, less commonly, IgG or IgA; they bind many toxins, carbohydrates and other common repetitive structures that are associated with various pathogens. The immunoglobulins may fix complement or bind to Fc receptors, thus permitting the construction of immune complexes for presentation before the development of acquired immunity.

Thus, a division of labor may emerge in intact antigen presentation. DC subsets may present antigen to naïve B cells, whereas the distinct lineage of FDCs may present antigens to activated B cells and evolving memory cells in the germinal centers. B cells may take the relayed antigen and present pro-cessed forms to T cells. Further study is required to establish the physiological importance of this chain of events now that the synaptic basis of a new link has been highlighted by Neuberger and colleagues2. Appre-ciating the immunological relay race may provide clues to enhancing the collaboration between DCs, B and T cells in vaccine development and provide approaches for strategically fumbling the antigenic baton, to help prevent or treat autoimmune diseases.