The polymeric antibodies secretory IgA (slgA) and IgM are important in immunity to influenza virus and this work was undertaken primarily to investigate the mechanisms by which they neutralized type A influenza virus (A/fowl plague virus/Rostock/34; H7N1) (FPV/R). Secondly, this thesis presents data on the quantitative aspects of neutralization by monoclonal and polyclonal IgG, focusing particularly on the number of antibody molecules binding to virus particles at minimum and maximum levels of neutralization. The attachment of FFV/R neutralized by slgA and IgM was completely blocked at 4°C. The large molecular structure of these antibodies suggests that steric hindrance may be the mechanism by which attachment is impaired. However at 25 and 37°C neutralizing slgA and IgM inhibited attachment of only 50% of the neutralized virus and none of this attached virus was internalized. It is inferred that IgM and slgA interfere with the endocytotic event responsible for internalization of virus. In contrast when neutralized by either IgG or monomeric IgA (obtained from the partial reduction of the above slgA), FPV/R attached to and penetrated into BHK-21 cells, with the genome accumulating in the nucleus, at a rate indistinguishable from that of non-neutralized virus. This behaviour was independent of temperature from 4 to 37°C. Thus IgG and monomeric IgA neutralized infectivity at a stage subsequent to penetration. Virus was neutralized by F(ab')2 and Fab’ fragments of monoclonal anti-HA IgG. Neither prevented the attachment, penetration, or the accumulation of viral genomes in the nucleus. Radiochemical data obtained on the binding of immunoglobulins (IgG, slgA) to the haemagglutinin (HA) of FFV/R in suspension suggests that the virus is saturated at approximately one four-chain immunoglobulin unit per HA spike. Two-step competition assays show that the binding of one molecule of monoclonal IgG to a HA spike prevents the binding of monoclonal antibodies to two different epitopes on that spike and also the binding of polyclonal antibody to the HA. These observations argue against direct epitope-epitope competition and for a steric or allosteric block on the binding of further IgG to the HA spike by prebound IgG. The kinetics of neutralization of FPV/R at 4°C with minimal concentrations of monoclonal IgG suggested that 3 IgG molecules were required to neutralize an infectious virus particle. The discrepancy between this and radiochemical data indicating that least 50 antibodies per virus particle are required for neutralization i3 reconciled by a theory that neutralization only occurs when antibody binds to certain "neutralization relevant" HA spikes, which are in the minority, and differ from "neutralization irrelevant" spikes only in their interaction with the viral core.

The potent cytotoxin ricin, obtained from the seeds of the castor oil plant Ricinus communis, is composed of two polypeptide subunits linked by a single disulphide bond. The binding of this molecule to the surface of eukaryotic cells is mediated via the sugar-binding activity of the B subunit. The exact nature of the ricin receptor(s) on the cell surface is unclear, but is most probably some glycoprotein or glycolipid containing exposed galactosyl residues. The ricin molecule becomes internalized by the cell and, by an unclear mechanism, ricin A chain escapes its endocytic vesicle entering the cell cytoplasm where it enzymatically and irreversibly inhibits eukaryotic ribosomes, bringing about the cessation of protein synthesis. The toxicity of ricin A chain-containing immunotoxins can in many model systems be enhanced by the subsequent addition of ricin B chain. This apparent ricin B chain-mediated potentiation of cytotoxicity suggests that this subunit has some role in the mechanism(s) facilitating the entry of ricin A chain into the cell cytosol. The work presented in this thesis has attempted to examine the potential of ricin B chain as a carrier of proteins into cells other than ricin A chain. In this example ricin A chain has been replaced with the type I ribosome inactivating protein, gelonin. Further to these studies, preliminary work considering the possible Importance of the hydrophobic C-terminus of ricin A chain in the translocation events, has been presented. To date this work has demonstrated that it is possible to delete at least 30 amino acid residues from the C-terminus of the A chain and retain full ribosomal inactivation activity as judged by in vitro analysis. This truncated form of ricin A chain has been expressed in an E. coli expression system and a soluble and active recombinant protein has been partially purified. The implications of this work and possible future analysis of this mutant polypeptide have also been considered.

Infiltrins, or pathogen-secreted host nucleus infiltrating proteins, are potential new targets for the development of more efficient vaccines against helminthic parasites. The archetypal infiltrin is IL-4 inducing principle from S. mansoni eggs (IPSE/alpha-1, also known as SmIPSE), a glycoprotein secreted by Schistosoma mansoni eggs, and characterised by the simultaneous presence of a classical secretory signal and a nuclear localisation signal. Within minutes following uptake by mammalian host cells, IP SE/alpha-1 translocates to their nucleus and binds to Deoxyribonucleic acid. This suggests that infiltrins, by acting as transcription factors, might play a central role in controlling the host-parasite relationship at the molecular level. Together with their secretory status, this role makes infiltrins interesting targets for vaccination. In this study, similar properties were demonstrated for nuclear localisation signal of ShIPSE03 (125-SKRRRKY-131) located between amino acid position 125 and 131, that plays a necessary and sufficient role in the process of transferring ShIPSE03 and heterologous GFP proteins into the nuclei of host cells. Similarly, a combination of online bioinformatics tools was used to predict the putative nuclear localisation signal motif of Smk5 (256-ELKRRVE-262) from S. mansoni eggs, and FhGST-si (202-LKKRAKT-208) and FhH2A (35-IHRHLKT-41) from Fasciola hepatica. To verify the predicted NLSs, putative infiltrins were used to generate a series of truncated constructs fused with Aequorea coerulescens green fluorescent protein-1 (AcGFP1), which were transfected into mammalian cells. Nuclear localisation of fluorescence confirmed the existence of a single, signal at the C-terminal in ShIPSE03, FhGST-si, and Smk5, and at the N-terminal in FhH2A. The predicted NLS motifs in ShIPSE03 (125-SKRRRKY-131), Smk5 (256-ELKRRVE-262), FhGST-si (202-LKKRAKT-208) and FhH2A (35-IHRHLKT-41) inserted into Tetra-enhanced green fluorescent protein (EGFP), but not corresponding alanine NLS mutants, redirected the encoded ~100 kDa protein entirely to the nucleus. Use of an IPSE-specific monoclonal antibody or an anti-His antibody showed that wild-type recombinant ShIPSE03, Smk5, FhGST-si and FhH2A added exogenously to HTB-9 or Huh7 cells, fully translocated to the nucleus, whereas the alanine NLS mutant remained in the cytoplasm. Overall, the existence of infiltrins in S. haematobium and F. hepatica suggests that infiltrins may represent a more general regulatory principle operating within parasitic trematodes. In terms of the function of IPSE/alpha-1, quantitative real-time polymerase chain reaction data indicated that an increase in Alanine-Transaminase (ALT) activity measured after 72 hours could be a result of increased gene expression after IPSE/alpha-1 nuclear translocation, rather than a true reflection of hepatotoxicity. According to Transepithelial Electrical Resistance (TEER) measurements in electrically tight Caco-2 cells grown in transwell inserts, wild-type IPSE/alpha-1 may play an important role in the down- regulation of intestinal epithelial cell tight junction integrity, might encourage apoptotic mediators and inflammatory responses in intestinal epithelial cells, and impair the intestinal tight junction barrier by causing it to dysfunction. In addition, wild-type IPSE/alpha-1 was able to activate humanised basophil reporter cell lines, such as the RS-ATL8 and NFAT DsRed cell lines.

Periodontal disease (PD) and rheumatoid arthritis (RA) are multifactorial chronic inflammatory diseases with high prevalence among the global population. There is evidence of a bidirectional relationship between PD and RA, although the underlying mechanisms remain undefined. Both PD and RA are associated with a dysregulated immune response and citrullination, a post-translational modification of proteins catalysed by peptidylarginine deiminases (PADs). PADs, in particular PAD4, are involved in formation of neutrophil extracellular traps (NETs) and may play a role both in generating potential auto-antigens and in host defence against bacterial infections. RA onset is preceded by a breach of self-tolerance and presence of anti-citrullinated protein antibodies (ACPAs). These ACPA have also been found in PD patients. Porphyromonas gingivalis is a key pathogen in PD and uniquely among prokaryotes expresses a PAD enzyme (PPAD), which is also potential source of citrullinated self-antigens. One hypothesis linking PD and RA suggests that the combination of PPAD and PAD4 activity in an inflamed environment may predispose to autoimmunity to citrullinated proteins and generation of ACPAs. This project aimed to determine the effect of PAD4 activity in PD and RA disease progression. Using PAD4 deficient animals or wild type controls, PAD4 was confirmed to be essential for NETs formation as bone marrow derived neutrophils from PADi4 knockout (KO) mice were unable to generate NETs in vitro. Experimental PD was initiated by oral infection with P. gingivalis and animals demonstrated a robust antibody response to P. gingivalis. However, there was limited evidence of bone loss in the animals, possibly due to inherent resistance in the strain. The immune response to P. gingivalis appeared unaffected by absence of PAD4, implying that NETS do not play a substantial role in the response to oral infection in this system. In experimental arthritis (EA) models, inflammation in EA was greater in absence of PAD4. Further investigation of the underlying mechanisms of PAD4 modulation of inflammation showed no direct impact in the innate response mediated by neutrophils, but confirmed a sexually dimorphic behaviour in PAD4 regulation of T-cell mediated inflammation. Pharmacological inhibition of PAD4 has been proposed and trialled as an RA therapy. These data suggest that PAD4 may impart subtle modulations on inflammation, which may impact on the outcome of such intervention.

The immune system is a complex network of cells and molecules working together with the purpose of fending off potentially harmful pathogens. CD4+ T cells take key roles within this network by orchestrating a multitude of its players. They recognise pathogen or self-derived peptides (p) bound to molecules of the major histocompatibility class II (MHC-II) through their T cell receptor (TCR). Cytokines secreted in response to recognition aid antibody production and cytotoxic T cell activity, both critical for anti-viral immunity. In this thesis, TCR/pMHC-II interactions were investigated using a range of functional and molecular approaches in order to gain valuable insight into the mechanisms underlying successful antigen recognition. To aid these investigations, a versatile, insect cell based expression system for HLA-DR1 was successfully implemented to generate soluble protein for use in multimer stainings and biophysical assays. Two HLA-DR1 restricted peptides encoded within influenza heamagglutinin (HA) were confirmed as being highly conserved making them ideal targets for vaccine development and allowing identification of influenza specific CD4+ T cells. Furthermore, the various roles of peptide flanking residues (PFR) were investigated using two experimental models. In a HA derived peptide, C-terminal PFR proved essential for peptide binding stability to HLA-DR1 and in consequence, CD4+ T cell activation. Clonotyping of CD4+ T cells grown against peptides of varying PFR lengths showed that TCR gene selection was heavily influenced by PFR. A HIV gag24 derived peptide displaying an unusual secondary structure within its N-terminal PFR gave further insight into how seemingly small modifications to PFR can have wide reaching impact on CD4+ T cell activation. Both studies highlighted the need for more in depths investigations into this emerging field and the wide reaching impacts of this inherent feature of MHC-II restricted peptides. Overall, the results in this thesis added novel insight into the mechanisms underlying TCR/pMHC-II interactions.

Staphylococcal protein A (SpA) produced by Staphylococcus aureus, binds Immunoglobulin G (IgG), forming SpA-IgG immune complexes (SIC), at a ratio of 2 SpA molecules to 4 IgG molecules. In vitro, SIC have been found to inhibit human osteoclastogenesis, and reduce proinflammatory cytokine secretion by human macrophages. The targeting of both osteoclasts and macrophages together may be important in diseases such as rheumatoid arthritis (RA), where inflammation and bone erosion both play roles in disease progression; resulting in considerable pain and disability. SpA is currently in clinical trials, however, as SpA forms heterogeneous IgG complexes, SpA therapy is not optimal. The aim of this project was to produce human recombinant Fc (Fragment crystallisable) constructs using different IgG subtype(s), thus optimising the immunomodulatory properties exhibited by SIC. This could lead to the development of new therapeutics, and will also give insight into Fc-mediated inhibitory mechanisms. In this thesis, 13 human Fc constructs (5 tetramers, 5 dimers and 3 monomers) were produced in total. The methods used to generate the human Fc constructs enabled their successful production for use in in vitro cultures. However, when investigating the structure of the IgG3 tetramer (TG3), IgG3 dimer (DG3) and IgG3 monomer (MG3), it was found that TG3 and DG3 do not appear to be polymerising in the manner that was predicted. The IgG1 tetramer (TG1) and TG3 tetramer, alongside SIC, were found to have the ability to significantly inhibit osteoclast differentiation, and the same molarities of dimers and monomers subsequently displayed reduced levels of inhibition. In studies performed using TG3, DG3, MG3 and SIC: TG3 and SIC were both shown to significantly inhibit RANK transcript levels, although they did not significantly affect the levels of RANK protein expression on the cell surface. However, the level of RANKL-mediated phosphorylated p38 (pp38) was reduced across all donors. Interestingly, TG3 and DG3 significantly reduced cell surface expression of CD115 (c-Fms; colony stimulating factor 1 receptor (CSF1R); macrophage colony-stimulating factor receptor (M-CSFR)). SIC and TG3 were able to significantly reduce transcript levels of DCSTAMP, OCSTAMP, cathepsin K and MMP-9 at certain time points, and TG3 was also shown to significantly reduce NFATc1 transcript levels. TG3, DG3, MG3 and higher amounts of SIC all displayed binding to the surface of cells, although this was only significant for DG3 and SIC. The mechanisms of action of the human Fc constructs and SIC have not been fully elucidated, and further work is required. However, the data in this thesis suggests that these novel human Fc constructs and SIC can exert their Fc-mediated inhibitory effects by affecting pathways essential for osteoclast differentiation and function.

Rotaviruses are important human and veterinary pathogens and are responsible for some 1-2 million human deaths per annum, worldwide. Conventional vaccine strategies for this pathogen have, on the whole been unsuccessful. Therefore, a detailed and comprehensive understanding of the immune response to rotaviruses, particularly at the cell mediated level is being sought, such that successful vaccines can be generated. A lymphocyte proliferation assay system has been developed for examining the Th cell response to the bovine rotavirus (BRV), UKtc. Splenocytes from adult BALB/c mice, orally inoculated with infectious BRV(UKtc) proliferated in response to in vitro stimulation with purified BRV(UKtc) particles. Proliferation was (i)detected at 4 days and 8 days after primary oral inoculation, (ii)not detected in uninoculated animals, (iii)specific to the priming virus and (iv)eliminated by NH4CI treatment of the spleen cells. Splenocytes from animals challenged by both oral and intra-peritoneal (i.p.) routes, were more efficiently stimulated by double-shelled BRV(UKtc) particles than single-shelled particles. Proliferation was found to be mediated by both Thy-1+,CD8−,CD4+ cells (i.e. Th cells) and Thy-1+,CD8+,CD4− cells (i.e. cytotoxic T-cells) but was dependent on Thy-1+,CD8−,CD4+ cells, when mice were inoculated by either the oral or i.p. routes. A greater proportion of the splenocyte proliferative response was found to be due to Thy-1+,CD8−,CD4+ cells when the animals were inoculated by the i.p. route. Virus replication in the intestinal tract was not required for a splenocyte proliferative response to be detected and the splenocyte response was long lived. For example, significant proliferation to both double and single-shelled forms of BRV(UKtc) was detected at 144 days (oral inoculation) and 224 days (i.p. inoculation), after a single dose of virus. BRV(UKtc) stimulated splenocytes secreted interferon-gamma, upon activation but no significant differences in titer were present between cells stimulated with double-shelled virus or single-shelled virus, in contrast to the [3H]thymidine incorporation results. Cross-challenge experiments with the porcine rotavirus OSU showed that cross-reactivity existed at the T-cell level. However, memory splenocyte proliferative responses to this strain were not long lived following oral inoculation with BRV(UKtc). The response of mesenteric lymph node cells to in vitro challenge with rotavirus was also studied at various times post oral inoculation. Of importance was the finding that proliferative responses to rotavirus were not present in mesenteric lymph node cells at 63 days post oral inoculation.

Lymph node cells from mice which have undergone primary cutaneous sensitization (responder cells) were cultured with either in vitro haptenated cells or hapten-bearing dendritic cells from hapten sensitized mice. By utilising the fluorescent hapten FITC and flow cytometry the stimulator cells hapten status was established and related to any enhancement in responder cell proliferation. In vitro haptenation failed to generate immunogenic hapten presenting cells but rather, hapten-coated cells whose stimulatory activity, while hapten-specific, was dependent on a silica-sensitive cell within the 'responder' population. The inadequacy of in vitro haptenated cells as a model for hapten presentation was thus established and a role for endogenous hapten processing cells is proposed. Hapten-bearing dendritic cells from hapten-sensitized mice were used as stimulators within the proliferation assay. Such cells were prepared (on density gradients) from FITC-sensitized mice and characterised in terms of morphology and flow cytometric parameters: these results correlated with a marked stimulatory activity for responder cells. Avoidance of potent sensitizing regimes enabled the isolation of dendritic cell-enriched fractions with hapten-specific stimulatory activity. Significantly, this activity could not be created by in vitro haptenation of naive dendritic cells. I concluded that the stimulatory activity of in vivo 'haptenated' dendritic cells within my proliferation assay was a good model for hapten presentation in vitro. Finally I examined the ability of in vivo administered, partially purified IFN gamma preparations (and relevant controls) to modulate cutaneous sensitization and in particular the generation of immunogenic dendritic cells. While not all the effects detected were IFN gamma specific, the changes measured in the hapten status and resulting stimulatory activity of dendritic cell-enriched fractions were so. Proposals aré made as to how antigenically foreign proteins and lympnokines, including IFN gamma, may regulate the role of dendritic cells in cutaneous sensitization.

Duffy-binding-like domains are present in two potential malaria vaccine candidates. Located on the merozoite surface, MSPDBL1 and MSPDBL2 have been implicated in erythrocyte invasion and identified as targets of natural immunity. Merozoite DBL domains have been shown to bind the Fc region of natural IgM. This is characteristic of several PfEMP1s, and is also well documented in bacteria, viruses and other parasites, where it is thought to prevent specific binding of the more deadly IgG antibodies. We have developed a mammalian expression system to produce merozoite DBL domains as Fc fusion proteins, facilitating investigation into their adhesive properties. Fc-fusion proteins are composed of the Fc region of IgG fused to a peptide and are a rapidly expanding field of bio-engineering. They have been successful in drug delivery due to their ability to increase serum half-life of the fused protein by the interaction of the IgG Fc with the neonatal Fc receptor (FcRn). Engineering of the Fc scaffold has shown improved receptor binding, allowing cross-linking of Fc receptors for improved vaccine design. The expression of homodimeric DBL-Fc fusions is di cult, evidenced by incorrect folding and low protein yield. A flexible ,extended hinge region was designed to increase the distance between the Fc and the fused DBL domain, and improved protein folding and IgM binding. Further work may optimise this hinge region for the development of malarial vaccines, or therapeutics for IgM-mediated diseases. The structural analysis of all known IgM-binding DBL domains and residues on the merozoite DBL surface predict the involvement of helix 2a in IgM binding. This contradicts a recent homology model of the IgM-binding interaction, and suggests that the model needs revision. An improved DBL-Fc fusion could be used to identify critical binding residues located in this helix using the more focused approach of site-directed mutagenesis.

T helper 2 (Th2) cell responses typify the immune response to parasitic organisms, which frequently invade the intestine. Dendritic cells (DCs) are considered vital for the induction of Th2 responses as they present parasite- derived antigens to naive T cells in draining lymph nodes. However, the identities of the DC populations responsible for priming Th2 cells in the intestine are still unclear. We developed an experimental immunization protocol to deliver Schistosoma mansoni eggs into the intestine. During live infection by the parasite, these eggs cause intestinal damage, granuloma formation, tissue fibrosis and strong type 2 immune responses. Many aspects of type 2 immunity are controlled by the transcription factor IRF4 and we observed that intestinal Th2 responses against Schistosoma mansoni eggs did not develop in the draining lymph nodes in the absence of IRF4+ DCs. IRF4f/f CD11c-cre positive mice had fewer CD11b-expessing migrating DCs, and fewer parasite antigen-carrying DCs were present in the mesenteric lymph nodes (MLNs) draining the small intestine and colon. However, transfer of antigen-loaded IRF4-deficient DCs directly into the MLN revealed that these cells could induce antigen-specific Th2 responses, suggesting that IRF4 controlled the migration of CD11b-expessing DCs rather than their Th2 inducing capacity. Furthermore, immature DCs from the intestinal lamina propria, and semi-mature DCs from lymph were sufficient to prime antigen-specific Th2 responses against egg antigens when transferred into naive recipient mice. This induction was dependent on MHCII expression but not on the production of IL-4 by the transferred DCs, indicating that conventional intestinal DCs are fully capable of inducing Th2 responses against S. mansoni egg antigens upon transfer. Further analysis of migratory small intestinal and colonic lymph DCs revealed that distinct subsets of CD11b-expressing DCs were sufficient for the induction of Th2 responses in the small intestine and colon. CD11b+CD103+ DCs transported parasite antigen from the small intestine, whereas CD11b+CD103- DCs performed this role in the colon. Of note, these same small intestinal and colonic DC subsets were also the populations that were most efficient at priming antigen-specific Th2 responses in vivo. Thus, we have not only identified that IRF4-dependent CD11b-expressing DCs are specialized to drive Th2 responses in the intestine but have also revealed that different DC subsets promote Th2 responses in the small intestine and colon. These findings not only advance our knowledge of intestinal Th2 responses against parasite antigens but also reveal a hitherto unappreciated functional heterogeneity among intestinal DCs, which could also be relevant for other tissue- specific intestinal conditions like Crohn’s disease, ulcerative colitis and celiac disease.