The presentation of multivalent antigens to CD4+ T cells

Abstract

Uptake of antigen for processing and presentation on MHC class II molecules is required for initiation of the adaptive immune response. The professional antigen presenting cells (APCs) - dendritic cells (DCs), macrophages and B cells - are responsible for this process. Many of the antigens encountered on the surface of foreign microorganisms are multivalent, such as viral capsid proteins. A great many more are rendered pseudo-multivalent by being held in an array on the cell surface - this would apply to any cell surface proteins. Despite this, antigen presentation is frequently modelled using monovalent soluble proteins such as hen egg lysozyme (HEL) or ovalbumin (OVA), which are unlikely to be representative of true antigens. We therefore investigated whether multivalent constructs of these model antigens were presented by APCs differently than their monovalent counterparts. Chapter 3 describes the creation and validation of multivalent model antigens. A range of different methods of homobifunctional chemical crosslinking were trialed, but most either destroyed the ability of the protein to stimulate CD4+ T cells or were unable to produce a product that was both multivalent and stable. These problems were circumvented by using the heterobifunctional crosslinker succinimidyl 4-(N-maleimidomethyl)-cyclohexane carboxylate (SMCC) in conjunction with N-succinimidyl-s-acetyl-thioacetate (SATA) to produce a multivalent construct made up of both HEL and OVA. Since it could be argued that any novel properties of this antigen could be specific to this construct, we employed other methods of creating multivalent antigens. Sulfo-NHC-LC-biotin was used to biotinylated OVA, which could then be assembled into multivalent complexes with the addition of avidin. This approach provided the additional advantage of variability, as varying the amount of avidin used in the assembly process altered final valency of the conjugate. Finally, peptides were synthesised with a C-terminal biotin and 0, 1 or 4 repeats of a DNP-tetrapeptide at the N-terminus, which could be assembled into 0-, 4-, or 16-valent constructs with the addition of avidin. These constructs allowed us to investigate the binding of antigens of defined valency to APCs, particularly B cells, although complications with antigen processing made these unsuitable for studying antigen presentation. The immunological properties of the different multivalent antigens were investigated in Chapter 4. All three constructs were efficiently captured by substantial subpopulations (~5%) of polyclonal B cells via their BCR. We demonstrated that different B cell subpopulations bound the different multivalent antigens, and that transgenic B cells expressing a BCR of irrelevant specificity were unable to bind the multivalent antigens. Additionally, both the HEL-OVA conjugate and the OVA-biotin-avidin conjugate were able to stimulate antigen-specific CD4+ T cell division at 10-100 fold lower concentrations than their monovalent counterparts, both in vitro and in vivo. This effect was dramatically reduced in B cell-deficient mice, indicating that antigen binding by B cells and enhanced CD4+ T cell proliferation were related. Finally, both binding to the BCR and CD4+ T cell proliferation were found to increase with antigens of increasing valency. In Chapter 5 we investigated the mechanism underlying the enhanced CD4+ T cell response to multivalent antigens. Interestingly, although polyclonal B cells were effective at capturing multivalent antigens via their BCR this interaction did not result in B cell activation, making them unable to stimulate CD4+ T cells directly. Instead, the enhanced presentation required cooperation between B cells and DCs. We determined that optimal CD4+ T cell proliferation required the B cells, but not the DCs, to have been exposed to antigen, and the DCs, but not the B cells, to express MHC class II molecules histocompatible with the CD4+ T cells. This indicates that multivalent antigen is captured by B cells and transferred to DCs for presentation. Furthermore, heat-killed B cells were unable to act as antigen donors, whereas apoptosis resistant bcl-2 transgenic B cells were effective donors, demonstrating that the antigen transfer is not simply a result of DC phagocytosis of apoptotic/necrotic B cells. Finally, we used fluorescently labelled multivalent antigen to demonstrate that intact antigen is indeed transferred from B cells to DCs, a continual transfer of surface molecules from naive B cells to a subset of DCs also being observed. Chapter 6 presents a general discussion of the research findings presented in this thesis. B cells have generally been thought to be irrelevant in the initiation of CD4+ T cell responses due to the scarcity of cells with a BCR of sufficient affinity to bind the antigen. We concluded that this is a flawed assumption resulting from the use of monovalent proteins as model antigens. Unlike monovalent antigens, multivalent antigens are efficiently captured by a substantial subpopulation of B cells, and presented, via DCs, to CD4+ T cells. Thus, our use of more appropriate model antigens has revealed a previously unrecognised antigen-capturing role for B cells in the initiation of CD4+ T cell responses. Show more