Search

The socio-economic and geographical distribution of malaria overlaps with that of many parasitic helminths and in these areas co-infections are common. Co-infection with helminths can influence disease outcome causing either exacerbation or amelioration of malaria. Understanding the complex host-parasite interactions that lead to these different disease outcomes is important for the success of control programmes aimed at these parasites. The immune system has evolved diverse types of response (e.g. T-helper 1 (Th1) and T-helper 2 (Th2)) to efficiently combat infection with ‘microparasites’ and helminths respectively. When faced with co-infection however, the need for the host to multitask means it must manage these counter-regulatory responses. In this study a murine model of malaria-hookworm (Plasmodium chabaudi- Nippostrongylus brasiliensis) co-infection was utilised to investigate how changes in T-helper bias affect malaria disease outcome. Antibody isotypes were used as indicators of Th1/Th2 bias and revealed that helminth co-infection reduced the malaria-specific Th1 response. Counter-intuitively this resulted in ‘protection’ from malaria with co-infected mice having reduced peak P. chabaudi parasitaemia and suffering less severe anaemia. In addition to providing a measure of Th1/Th2 bias, analysis of antibody responses revealed the occurrence of cross-reactive antibodies. The potential for these crossreactive antibodies to influence disease outcome was investigated but in this murine model resource-mediated mechanisms of parasite regulation appear to be responsible for the ‘protection’ that co-infection affords. The question of why cross-reactive antibodies are produced has important immunological and ecological implications. Cross-reactive responses may arise through some physiological constraint on the immune mechanisms that usually result in antibody-specificity. However experiments designed to investigate if the specificity of antibodies is constrained by availability of antigen suggest that this is not the case in the model system used here. There is also the possibility that production of cross-reactive antibodies represents an evolutionary optimal strategy for a host faced with unpredictable exposure to a variety of parasites. However a major finding of this study indicates these two taxonomically distinct parasite species share antigens, which in itself is crucial to understanding host-parasite interactions in a co-infection setting. The main findings of this thesis are relevant to co-infection studies in general and the implications for both evolutionary and applied biology are discussed.

Malaria is a major cause of illness and death in the world. Much progress has been made in the development of a vaccine against malaria, but there is still no formulation that can induce effective and long-lasting immunity in diverse populations at an affordable cost. Bacterial lipoproteins are potent stimulators of the innate and acquired arms of the immune system. They have been used as vaccines against bacterial diseases, and short lipopeptides have been shown to be potent adjuvants to a diverse range of vaccines against bacterial and viral diseases, and may also be useful in the development of anti-tumour vaccines. Here, the artificial lipidation of selected malaria antigens, MSP119 of Plasmodium chabaudi chabaudi AS and a multi-epitope string comprising mainly P.falciparum epitopes (SGC), is investigated. Direct lipid modification of the antigens was achieved by fusion of the protein coding sequence to a bacterial lipidation signal sequence that directs the post-translational addition of fatty acids to the protein. Antigens were also expressed as fusion proteins to a naturally occurring lipoprotein of Pseudomonas aeruginosa, L-OprI. The inclusion of L-OprI as an admixed adjuvant boosted the antibody response to MSP119, and although this response was not protective against a live blood-stage challenge, neither did it lead to increased parasitaemia and mortality. Immunisation of mice with the multi-epitope string SGC as a fusion to the L-OprI lipoprotein induced an antibody response against the SGC component of the molecule while immunisation with protein alone did not. This antibody response was absolutely dependent on covalent linkage of the L-OprI lipoprotein to SGC; the addition of LOprI as an admixed adjuvant did not induce an antibody response to SGC. SGC-specific IFNγ-secretion by whole spleen cells was induced by immunisation with protein, fusion protein or fusion lipoprotein forms of SGC, but there were no significant differences in the numbers of IFNγ-secreting cells between preparations. There are several technical obstacles that must be overcome before artificial lipoproteins can be produced on a large-scale, but efforts to overcome these obstacles would be worthwhile since lipoproteins appear to represent suitable candidates for further investigation as adjuvants to malaria vaccines.

Many believe that vaccines are the intervention with the greatest chance of reducing the worldwide burden of malaria. Irradiated sporozoites protect against subsequent malaria challenge, and a vaccine that mimics this stage might be effective. CD4 and CDS T cells are thought to be crucial in protection, but studies of T cell immunity in malaria exposed donors are hampered by widespread nonresponsiveness, to malaria antigens. Most studies of natural T cell immunity measured lymphoproliferation alone, but I found that if 3 T cell assays were employed simultaneously (proliferation, overnight IFN-y ELISPOT, and IFN-y ELISPOT after 14 days culture), that T cell reactivity to circumsporozoite protein (CS) was considerably higher than when using proliferation alone. Responses to individual epitopes failed to correlate over the 3 assays, suggesting that they detect different memory T cell subsets. I investigated the nature of responder cells in the 3 assays, and found ex- vivo ELISPOT responses were mediated by CCR7+ and cultured ELISPOT by CCR7+ cells. I also found a CD4+CD38+ T cell subset that suppressed lymphoproliferation, and might contribute to malaria immunosuppression. The immunodominant T cell epitope regions of CS are highly polymorphic, and I found that altered peptide ligand antagonism operated for variants of a CD4 T cell site. Less polymorphic antigens might be better vaccine candidates, one promising candidate being thrombospondin related adhesive protein (TRAP). I mapped novel CD4 T cell epitopes in TRAP, 16 of which were conserved. I found differences in TRAP T cell reactivity between East and West Africans, and adults and children, which is important since these populations will be recipients of future malaria vaccines. I investigated whether IFN-y ELISPOT responses to TRAP correlated with protection, since IFN-y is thought to be the mechanism by which T cells exert their protective effect, but no protection was found.

The genetic and immunological basis of canine atopic dermatitis (CAD), a hypersensitivity disease to environmental allergens in dogs, was investigated. The evaluation of the clinical records of 434 dogs belonging to 92 litters related to dogs with CAD showed that the disease was greatly dependent upon genetic influences with a heritability of 0.49. Using a highly specific IgE detection reagent (FcƐRl alpha-chain) Dermatophagoides pteronyssinus (DP) -specific IgE was measured in the serum of normal and atopic dogs by ELISA. Dogs with CAD had significantly more DP-specific IgE than non- atopic dogs (p<0.0001). Immunoblotting revealed that the serum IgE response to DP in CAD was dominantly to three polypeptides with apparent molecular weights of 34, 63 and 89-95kDa. The 89- 95kDa polypeptide has been characterised as a mite chitinase. Cellular immune responses to DP extract were measured in normal and atopic dogs by measuring the antigen-specific proliferation of peripheral blood mononuclear cells. Atopic dogs showed greater proliferative responses to DP than normal dogs, although there was large individual variation. Dogs were immunised with DP extract to help identify the approximate molecular weights of allergens containing the dominant T-cell epitopes. The lymphocyte proliferative responses to DP polypeptides purified by SDS-PAGE were greatest to proteins of between 60 and 90kDa, similar in size to those reacting with serum IgE in dogs with CAD. Clinical and laboratory examination of dogs with allergic conjunctivitis showed that this condition was associated with CAD. Histopathological examination of conjunctival tissue often revealed signs of allergic response. However, IgE was not detected in the tears of enough affected dogs for it to be diagnostically useful. Family genetic studies and selective breeding programmes can significantly reduce the numbers of animals suffering from CAD. This study has identified antigens which should be used as the basis of improved CAD diagnostic and therapeutic reagents.

The aims of this study were: 1. To isolate the specific autoantibody secreting B cells, by generating human anti-PDC-E2 monoclonal antibodies and precisely defining their specificities and functional properties. 2. To evaluate the immunoglubulin genes that encode these autoantibody structures, in order to explore the genetic basis of antigen recognition, and to further understand their origins in primary biliary cirrhosis. Five heterohybridomas secreting PDC-E2 specific monoclonal antibodies, closely resembling the hallmark autoantibodies of PBC, were generated from the peripheral lymphocytes of two female patients. Genetic analysis revealed no evidence of restricted immunoglobulin variable region gene usage, and therefore argued against the operation of a superantigen in the induction of antimitochondrial antibodies. Moreover, the pattern of heavy chain variable region (V_H) gene mutations suggested an antigen-driven selection of high affinity IgG autoantibodies, supporting a possible role for exogenous antigen in the pathogenesis of PBC. This may be an important determination of antigen recognition. In further experiments, the precise structure-activity relationship of a single PDC-E2 specific monoclonal antibody (PD2) was explored. The results showed that antigen recognition required co-operative recognition by the heavy and light chains and that the heavy chain CDR3 was pivotal. Somatic mutations made only a slight difference to recognition of whole PDC. Reactivity to recombinant antigens, by contrast, was dramatically altered, resulting in a complete shift in antigen specificity from the E1/E3 binding domain to the inner lipoyl domain. This novel finding may be significant to the pathogenesis of primary biliary cirrhosis.

The ability of members of the hsp70 family to bind to peptides in vivo and in vitro suggests that they may be involved in the processing of antigens for binding to Major Histocompatibility (MHC) class I and/or class n molecules. The aims of this thesis have been to provide evidence for the involvement of hsp70s in antigen processing and to characterise the binding of peptides by hspTOs by structural and functional studies. Firstly, the peptide-binding domains of two hsp70s, hsp70hom and PBP74, were expressed in isolation from the rest of the molecule for structure determination. Both of these hsp70s were implicated in antigen processing: hsp70hom in the class I pathway, due to its cytoplasmic localisation and constitutive expression, and the presence of its gene in the MHC; and PBP74 in the class n pathway because published work indicated that it was localised to endosomes and that antibodies against it inhibited antigen processing. The expression and purification of both peptide-binding domains was very successful, and one dimensional NMR experiments indicated that they were folded. However, it was not possible to determine their structures by NMR spectroscopy or X-ray crystallography because they aggregated in solution at high concentrations. Instead, the structure of the C-terminal region of hsp70hom, which includes its peptidebinding domain, was modelled based on the known structure of the equivalent portion of dnaK, the hsp70 of E.coli. The structure of hsp70hom is predicted to be very similar to that of dnaK, and modelling studies suggest that it is likely to bind peptides in a closely related fashion. The modelling of complexes between hsp70hom and two peptides suggest that the peptide-binding groove is very versatile, accounting for the broad peptide-binding specificity of hsp70s. The interactions of hsp70hom and PB74 with peptides were investigated using plate binding assays and isothermattitration calorimetry. A biotinylated peptide bound to the peptide-binding domain of hsp70hom, immobilised in plastic wells, with a Kd of <25 μM, which is within the range of Kds reported for other hsp70-peptide complexes (0.1-100 μM). In solution, isothermal titration calorimetry showed that the binding of peptides to the peptide-binding domains of hsp70hom and PBP74 was likely to be entropically rather than enthalpically driven, and, therefore, the interactions involved are likely to be predominantly hydrophobic. Secondly, PBP74, an hsp70 thought to be involved in the class II antigen processing pathway in endosomes, was localised by immunofluorescence microscopy. It was shown to be a mitochondrial protein, and is, therefore, unlikely to be involved in antigen processing. The presence of other members of the hsp70 family in lysosomes purified from a B cell line by Percoll density gradient centrifugation was investigated using antibodies that reacted with many Afferent members of the hsp70 family. No hsp70s were detected in these late endocytic compartments, even after heat shock or serum starvation. However, the presence of an hsp70 in endosomes, or of a member of this family not detected by the antibodies used, in lysosomes, cannot be ruled out. A third approach investigated the induction of the three hsp70 genes found in the MHC by four cytokines. The hsp70-l and hsp70-2 genes are induced at the mRNA level by IFN-γ and IL- 1, while TNF induces hsp70-2 alone. This data supports a role for the heat-inducible hsp70 in MHC class I antigen processing, as it appears to be coregulated with known members of this antigen processing pathway. The expression of hsp70hom was unaffected by any of the four cytokines examined. In addition, the mitochondrial hsp70 (which is not encoded in the MHC) appears to be induced by IFN-γ at the protein level. The research presented in this thesis provides a greater understanding of the peptide-binding properties of two hsp70s. Further work is necessary to show conclusively whether any of the hsp70s is involved in antigen processing.

In the course of the work presented in this thesis, the construction of solid matrix-antibody-antigen (SMAA) complexes as vaccines was further developed. In particular, it was demonstrated that it is feasible to assemble SMAA complexes using a short oligopeptide tag (Pk) attached to the C-termini of antigens and a Pk tag- specific mAb SV5-P-k. In order to facilitate the purification of recombinant proteins for immunization purposes, a second affinity tag was attached to the antigen N-termini. Initially, the N-terminal tag was 26-kDa-large thrombin-removable glutathione S-transferase (GST), which permitted first-step purification on immobilized glutathione. However, because of problems with protein insolubility and the proteolytical removal of GST from the hybrid proteins, the GST domain was substituted by an N-terminal 12-amino acid-long tag (His) containing an array of 6 histidines. The His tag was small and thus did not require removal prior to immunization, and allowed purification of His-linked proteins on a nickel-affinity column. Moreover, it was possible to preform nickel-affinity chromatography under protein denaturing conditions, which allowed purification of insoluble or aggregated proteins. In addition, novel prokaryotic expression vectors were constructed for a single-cloning-step addition of these N- and C-terminal tags to proteins of interest. These vectors were used to individually express all non-glycosylated products encoded by the simian immunodeficiency virus (SIV) in E. coli. The SIV envelope glycoprotein gp160 with the Pk tag attached to its C-terminus was expressed in insect cells and first-step purified on a lentile lectin column. Following the first purification step on either nickel or lentile columns, all SIV proteins were purified and successfully incorporated into SMAA complexes using anti-Pk tag mAb SV5-P-k. Thus, efficient purification protocols were developed, which purified recombinant proteins via two different affinity tags attached to their N- and C-termini and isolated predominantly full-size proteins. As a stage in achieving the goal of human multivalent vaccines, the SV5-P-k mAb was humanized and is currently being expressed in Chinese hamster ovary cells.

Human cytotoxic T lymphocytes (CTL) recognize antigen in the form of short peptide fragments presented by Human Leukocyte Antigen (HLA) class 1 molecules. The HLA B27-restricted CTL response to Influenza A virus is directed against a single nine amino acid epitope within the viral nucleoprotein NP383-391. Influenza A-specific CTL lines and clones, generated from the peripheral blood of three unrelated individuals, exhibited remarkable fine specificity for NP383-391, failing to recognize synthetic variants containing single amino acid substitutions at positions 1, 4, 7 or 8. The sequences of rearranged T cell receptor (TCR) alpha and beta chain genes showed remarkable conservation of variable and joining gene segment usage, and also of non germline-encoded B chain residues proposed to interact with peptide, even between clones from unrelated individuals. Correlation of fine specificity for peptide with TCR alpha chain usage suggested a possible orientation for the TCR/HLA/peptide ternary complex in this immune response. The finding of restricted TCR usage in the HLA B27-restricted CTL response to influenza A virus has implications for the study and potential therapy of HLA B27 associated autoimmune disease. The TCR heterodimer from one clone was expressed in a rat basophil cell line and shown to confer specificity for HLA B27 + NP383-391. Influenza A-specific CTL were also shown to recognize bacterial superantigens in an MHC class II dependent but unrestricted manner. A rapid screening assay for candidate superantigens was developed and used to identify a novel superantigen produced by Clostridium perfringens. This superantigen was found to have a unique specificity for human TCRβ chains.

The aim of the work described in this thesis was to prepare and characterise a library of monoclonal antibodies that recognise antigens present in the insulin secretory granule membrane. These were to be used in a study of the composition, biosynthesis and turnover of the secretory granule membrane.

Data from ¹H nuclear magnetic resonance studies on the Fv fragment of protein 315, a Dnp-binding BALB/c mouse lgA(λ2) myeloma protein, have been used to refine a predicted structure of the combining site of the protein. The Dnp-binding subsite in the modified structure is composed of the side chains of three aromatic amino acids Trp 93L, Tyr 34L and 34H. A fourth aromatic amino acid residue is close to the side chain-NH-CH 2 -group, this is Tyr 33 H - The antibody-hapten binding is a simple encounter process, which causes no extensive conformation change in the Fv fragment. A method for paramagnetic structural studies has been devised using Dnp derivatives with cllgophosphate side chains, which create a specific manganese binding site on the Fv fragment-hapten complex. The distances from the bound metal ion to the imidazole side chains of two of the three hlstldine residues of protein 315 have been determined. ¹H nuclear magnetic resonance has been used to study the histidine residues of the Fv fragment of protein 315. It has been shown that one of the three histidine residues (102H) is close to the combining site, but that this residue does not participate directly in binding haptens. 31 P nuclear magnetic resonance studies have shown the presence of a positively charged amino acid side chain near the entrance of the combining site of the Fv fragment. This residue has been identified as Arg 95L. The mode of binding of trini trophenyl derivatives to the Fv fragment has been studied by 1H nuclear magnetic resonance. It is concluded that these haptens, when bound to the Kv fragment, make contacts with the same amino acid side chains as Dnp derivatives.

A new monoclonal antibody against rat lymphocytes is described. This antibody, designated MRC-OX22, recognises an antigen found on all peripheral B cells, all suppressor/cytotoxic lymphocytes, but only on some helper cells, as defined by the W3/25 antibody. The antigen is found on a significant number of bone marrow cells, but only on 3% of thymocytes, which were located by immunoperoxidase staining at the cortico-medullary junction. The phenotypic division of the W3/25+ peripheral T cells, is mirrored functionally. By using the popliteal lymph node assay, and the induction of lethal graft-versus-host disease as indicators of alloreactivity, it is found that the alloreactive T cells have the phenotype MRC-OX22+ ,W3/25+. In contrast, using an adoptive transfer system to assess T cell help for B cell antibody production, the phenotype of the Thelper for B cells is shown to have the phenotype MRC-OX22-, W3/25+. Evidence is also presented that the phenotype of the Thelper cell for B cells is stable after priming and boosting. The phenotype of the memory B cell is shown to be MRC-OX22+, and the cell responsible for allogeneic suppression is also shown to be MRC-OX22+, using the adoptive transfer system for antibody formation. The significance of the functional heterogeneity of the W3/25+ subset is discussed, and the biochemical nature of the antigen recognised by MRC-OX22 is reviewed, and compared with possible homologues in mouse and man.

This year, more than 2.1 million people are likely to die from lung cancer, and half of these deaths will occur within the first year of diagnosis. Lung cancer represents one of the most aggressive variants of cancer and is in significant need of new therapies. Immunotherapy represents a promising area of development and has seen some significant breakthroughs in recent years. While these new developments have advanced the field significantly, they largely do not approach the problem from the angle of prevention - the only true way to eradicate disease. Cancer vaccines, however, have the potential of a therapeutic benefit as well as the possibility of developing immunity and ultimately prevention. In response to infection, the cells of our immune systems generate swift and aggressive responses to pathogens and generate protective immunity that can often last decades. Unfortunately, when our immune systems respond to cancer, they almost always fall short and fail to completely eradicate tumour cells or establish effective immunity. Mimicking a microbial infection in response to cancer, therefore, seems a plausible way to bridge this gap. This study sought to determine whether bacterial CpG DNA, a potent inducer of macrophage activation, could improve immune responses to lung cancer antigens, which may function as a unique membrane lysate cancer vaccine. Within this study RAW 264.7 and THP-1 macrophages were exposed to CpG ODN 7909 with increasing concentrations of membrane protein, harvested from the A549 lung cancer cell line. Statistically significant increases in the production and expression of nitric oxide, TNF-α and IL-6 as measured by ELISA and Real-time PCR, were observed in response to CpG-antigen formulations. These findings were further supported with qualitative analyses confirming the uptake of CpG DNA, the activation NF-κB and the increase of membrane bound MHC II molecules via confocal microscopy. These results together, demonstrated some evidence of a dose dependent relationship between A549 membrane antigen in the production of nitric oxide and TNF-α. Macrophages did not exhibit increases in the production or expression of human or murine IL-12 or MHC II. The next step for these findings, is to test this vaccine formulation in vivo with a prophylactic focus. Where CpG ODN and membrane antigen formulations can be more comprehensively evaluated within a functioning immune system, and to further the growing field of preventative immunotherapy.

Engineered T cells are proven cancer therapeutics capable of executing potent antigen-dependent cell killing, and have gained FDA-approval after unprecedented results against certain blood cancers. However, successful treatment of solid tumors has been limited in large part by the lack of targetable antigens that are not also expressed in healthy tissues, as exhibited by ON-target OFF-tumor cross-reactive fatal toxicities seen in various human trials. An advantage of cell-based therapies is that cells can “compute” and execute more sophisticated programs than traditional drugs. Combinatorial antigen recognition has recently been introduced to overcome ON-target OFF-tumor toxicity limitations associated with traditional single antigen-targeted T cell therapies. This dissertation focuses on developing T cell engineering and biomaterials toolkits to expand the capabilities of combinatorial antigen-sensing T cell therapies. First presented is a set of approaches that use synNotch receptors and other synthetic receptors to program T cells with a diverse array of novel combinatorial antigen sensing circuits. It is shown that these Boolean logic-gated synNotch T cells can be engineered to sense both extracellular and intracellular antigens, target up to 3 antigens, integrate positive and negative regulation, and incorporate bispecific receptors. These abilities could enable engineering therapeutic T cells with the ability to target truly tumor-specific antigen combination signatures identified via bioinformatic analysis of antigen expression in cancer versus healthy cells. Also presented is a system to engineer T cells to recognize an orthogonal cue on biocompatible particles and incorporate this detection into combinatorial antigen-sensing circuits to specifically kill antigen positive tumor cells only in the presence of the particles. Using particles to control therapeutic cell activity in the body has the potential to enhance safety by enabling dynamic, local user-control of engineered cells during treatment. Together, the technologies presented here facilitate the development of engineered T cells to safely target a broader range of cancers. The toolkits developed here can be used to engineer T cell therapies to target cancer cells with significantly improved precision by taking advantage of our modern capabilities in cancer informatics, cell engineering, and biomaterials.