Respiratory viral infections caused by Influenza A viruses (IAVs) are responsible for annual seasonal epidemics that cause mild, severe or fatal pulmonary disease and estimated to result in 243,000-640,000 deaths globally every year. Severe infections can result in lung tissue damage, acute respiratory distress syndrome (ARDS) as well as secondary bacterial pneumonia, all of which are detrimental to the host and can potentially lead to a fatal disease outcome. Many of those at high-risk of developing severe disease include children, the elderly and immunocompromised individuals. Although current influenza vaccines are the best way to combat the seasonal influenza epidemics, the vaccines need to be updated annually to match strains predicted to circulate in the upcoming influenza season. Frequent mutations in the virus can lead to strain mismatches, which can affect the protective efficacy of these vaccines and result in significant increases in infection rates, putting those who are most vulnerable at risk for developing severe disease. The aim of this PhD thesis was to investigate host factors and mechanisms associated with severe influenza disease to provide insights into identification and prediction of severe disease outcomes. Co-expression of surface CD38 and MHC-II on CD8+ T cells is recognized as a classical hallmark of activation in many acute and chronic viral infection settings. However, high and prolonged CD38+MHC-II+ expression can be also associated with severe and fatal disease outcomes. Our previous studies have found that the persistence of CD38+MHC-II+ CD8+ T cells is associated with impaired IFN-gamma production in H7N9-infected patients with fatal outcomes. In Chapter III, we sought to investigate the mechanisms underpinning the expression of CD38+MHC-II+ markers using wild type and transgenic mouse models of influenza infection. We provided evidence that CD38+MHC-II+ CD8+ T cells are recruited early to site of infection during severe influenza A virus infection and can be activated in an antigen-driven as well as bystander manner. Our adoptive transfer experiments also demonstrated that MHC-II molecules were not expressed intrinsically by CD38+CD8+ T cells, but acquired from antigen presenting cells by trogocytosis, and the CD38+MHC-II+ phenotype was crucial for optimizing T cell recall ability. In Chapter IV, we investigated disease signatures during the early phases of infection to identify features associated with clinical outcomes. While comparable viral titres were observed between H7N9 patients who survived or succumbed to infection, there was a trend of higher level of inflammation responses in those who died. Our transcriptome analyses of blood samples from fatal and recovered H7N9 patients found that the gene Olah, encoding for a fatty acid hydrolase, was expressed more than 80-fold higher in fatal H7N9 patients compared to those that recovered at early days after disease onset. We established an Olah-overexpressing A549 cell line, which allowed us to develop a qPCR platform to screen Olah expression in human tissues and cells. Our analyses of Olah distribution across different anatomical sites found Olah expression in human lung, spleen and PBMCs, with the most prominent expression detected in CD14+ cells within the human lungs. Subsequently in Chapter V, we generated and established OLAH knockout (OLAH KO) mice to further investigate the role of Olah during severe influenza A virus infection. Compared to wild-type mice, OLAH KO mice displayed milder disease symptoms, less body weight loss and were protected from severe and fatal disease outcomes following IAV infection. In addition, OLAH KO lungs exhibited lower viral titres, less tissue damage and had significantly reduced levels of inflammatory cytokines and infiltrating innate immune cells, indicating the importance of Olah in controlling early excessive host responses and disease severity. These effects were associated with reduced lipid droplets in virus-infected epithelial cells lining the bronchioles, thus defining a role for OLAH in driving lipid metabolism during IAV disease. Overall, the findings from this PhD thesis highlight the importance of CD38+MHC-II+ CD8+ T cells during severe influenza A virus infection and identified the mechanisms underlying severe IAV disease. We also identified Olah as a potential biomarker of severe disease and demonstrated its role in modulating early host immune responses via the formation of lipid droplets and defined its impact on disease outcomes. Overall, data generated in this PhD thesis advance our understanding of factors underlying fatal clinical outcomes and provide insights into the development of novel treatment strategies and biomarkers against severe influenza disease.

Mucosal-associated invariant T (MAIT) cells are a subset of innate-like alpha/beta T cells that recognize riboflavin metabolites presented by the monomorphic major histocompatibility complex (MHC) class I related protein-1 (MR1). The most potent antigen known to date is 5-(2-oxopropylidineamino)-6-D- ribitylaminouracil (5-OP-RU). MAIT cells are abundant in mucosal tissues and blood in humans. Upon bacterial infection, MAIT cells expand rapidly, with production of cytokines, including interferon-gamma (IFN gamma), tumor necrosis factor (TNF), granulocyte macrophage colony stimulating factor (GM-CSF) and interleukin-17 (IL-17), and cytotoxic granzymes. Their phenotype indicates that they play an important role in immunity. Previous studies showed a protective role in local infections, involving a single organ or tissue, but the mechanisms of MAIT cell-mediated protection in systemic infections are not fully understood. This study aimed to elucidate MAIT cell activation, protective role in primary infection and potential for a MAIT cell-based systemic vaccination to protect against Francisella tularensis live vaccine strain (LVS) and Legionella longbeachae. F. tularensis is a gram-negative intracellular bacterium which can cause systemic infection, in mice and humans. In this study, F. tularensis LVS was used to induce systemic infection in C57BL/6 (wild type) mice and in Mr1 -/- mice, which lack MAIT cells. A combination of CpGcombo (fused oligos for CpG-B and CpG-P) and synthetic 5-OP-RU antigen was used to vaccinate mice. Bacterial load, survival rate, and MAIT cell response kinetics and post-infection cytokine profiles were examined. MAIT cells expanded systemically and showed Th1-like cellular profile after infection with F. tularensis LVS. In several organs, C57BL/6 mice showed better control of bacterial burden compared with Mr1 -/- mice. Vaccination of MAIT cells with CpGcombo plus 5-OP-RU, but not CpGcombo alone, protected mice from infections by an otherwise lethal dose of F. tularensis LVS and L. longbeachae. Thus, MAIT cells displayed a protective role against systemic infections and potential to be boosted to protect against local and systemic infections. This study also showed that post infection, MAIT and non-MAIT alpha/beta-T cells manifest different contraction kinetics, indicating that these two groups could be regulated by different viability mechanisms. Specifically, the in vivo data illustrated that loss of receptor-interacting serine/threonine-protein kinase 3 (RIPK3, a key regulator of apoptosis and necroptosis), but not mixed-lineage kinase domain like pseudokinase (MLKL, a signaling molecule involved in necroptosis), preferentially increased MAIT cell abundance in a cell-intrinsic manner. In summary, the results in this thesis demonstrated that MAIT cells are critical in immune protection against systemic F. tularensis LVS infection, and are long lived with a differently regulated cell death and survival. These findings may inform the future development of vaccination strategies targeting MAIT cells.

Vaccination has been an incredibly successful public health intervention, saving the lives of 2-3 million people each year. Despite this success, we still lack effective vaccines for many infectious diseases including HIV, tuberculosis and malaria. Nanoparticles (ordered structures within the range of 10-1000nm) have great potential to supplement traditional vaccines based upon pathogen subunits, or killed or attenuated microorganisms, as demonstrated by the successful licensure of virus-like particle vaccines for human papillomavirus and liposomal mRNA vaccines for SARS-CoV2. However, the immunological mechanisms that explain the potent immunity of nanoparticle vaccines and the factors dictating their interaction with the immune system are poorly defined. This thesis studies how nanoparticle characteristics affect their interaction with the immune system with a view to improving vaccine strategies. First, the contribution of the protein corona on the association of engineered nanoparticles with primary human blood cells was assessed. The association of high protein binding (high-fouling) mesoporous silica (MS) particles and low-fouling zwitterionic poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC) particles with human white blood cells was assessed by flow cytometry in the presence or absence of plasma proteins. The effect of precoating nanoparticles with serum albumin, IgG and complement protein C1q was also assessed. The differential association of low and high-fouling nanoparticles was found to be largely a consequence of the de novo formed, not pre-adsorbed, biomolecular corona. Specifically, an enrichment of complement proteins within the corona resulted in an increased association with B cells. Second, the immune mechanisms that give rise to the improved immunogenicity of a prototypic nanoparticle vaccine were investigated. Humoral immune responses to a self-assembling protein nanoparticle vaccine for influenza (HA-ferritin) were contrasted to a subunit influenza vaccine (soluble HA) in mouse and non-human primate models. Antibody titres and protective efficacy of the vaccines were compared followed by a detailed study of lymph node germinal centre B cell and T follicular helper cell responses. Vaccination of C57BL/6 mice with HA-ferritin nanoparticles elicited higher serum IgG titres and greater protection against experimental influenza challenge compared with soluble HA vaccination. Within the antigen-draining lymph nodes, germinal centre reactions were expanded and persistent following HA-ferritin vaccination. This augmented humoral immunity was not driven by ferritin-specific T follicular helper cells but rather driven by expanded antigen colocalization with follicular dendritic cells. However, this immune enhancement did not translate from mice to pigtail macaques where antibody titres and lymph node immunity following HA-ferritin nanoparticle vaccination were comparable to soluble HA protein vaccination. And thirdly, we explored innate immune activation by HA-ferritin and soluble HA in mice. This was achieved through in vitro assessment of antigen glycosylation and complement activation and in vivo through serum IgM titres and cell trafficking to the lymph nodes following vaccination. HA-ferritin vaccination of mice was found to elicit an early enhancement of antigen-specific serum IgM however in vitro complement activation was not detected. Trafficking of immune cells to the lymph nodes was found to be influenced by antigen glycan composition in conjunction with purification methods. The findings of this thesis suggest that nanoparticle interaction with the immune system is driven by the complex interplay of nanoparticle physiochemical properties, antigen glycosylation, corona formation and pattern-recognition receptors of innate immune cells. Further improvements in understanding the relationship between these features and how they may differ between animal species will speed the rational design of next-generation nanoparticle vaccines against diverse pathogens.

Tuberculosis (TB) elimination is now an aspirational goal in many low-incidence settings,1 born from an adaptation of the World Health Organization End TB Strategy, which states that low-incidence countries must progress towards eliminating TB as a public health problem. The TB incidence in Australia declined significantly during the twentieth century, but progress towards TB elimination has been influenced by high levels of migration from high-incidence countries. Most TB cases now occur among overseas born residents, and are attributed to reactivation of latent TB infection (LTBI) acquired overseas. LTBI is asymptomatic and not infectious, but those with LTBI can be treated to reduce their future risk of TB reactivation. However, there is uncertainty regarding the prevalence and distribution of LTBI in the Australian population and the risk that LTBI poses for future TB reactivation. Therefore, the effectiveness of LTBI screening and treatment strategies is also uncertain. The aims of this work were to: evaluate aspects of existing TB management in Australia; better understand and define the risk of LTBI and TB reactivation; and determine whether additional LTBI screening and treatment strategies for overseas-born residents could be cost-effective and advance Australia towards TB elimination. The thesis begins by reviewing current TB management in Victoria in relation to two factors identified as having an important impact on patient care: the private sector and gender. Observed disparities in the care received in the private versus public sector appeared to be primarily due to differing case-mix, and the small disparities in TB management and outcomes seen by gender were consistent with the possibility that males present with more severe disease, rather than any inequity in TB management. With this knowledge, Chapter 3 begins to explore the potential for additional LTBI screening and treatment strategies. Global annual risk of TB infection estimates were applied to Australian census data to estimate the LTBI prevalence in Australia, and identify those groups at the highest risk of LTBI. Two chapters are then devoted to improving our understanding of the risk of TB reactivation in the years and decades following infection and quantifying this risk. Chapter 4 presents a systematic review of the existing evidence on this topic and revealed that significant gaps remain in our understanding of reactivation risks in the years after infection. Chapter 5 combines the LTBI estimates from Chapter 3 with Australian TB case data to reveal that reactivation risks in overseas-born Australians largely decrease with increasing time from migration, with additional possible increases in risk during youth and in the elderly. In Chapter 6 the findings from previous chapters are applied in a cost-effectiveness analysis of additional LTBI screening and treatment strategies in recent Australian migrants. Modelled strategies prevented between 2.3 and 16.9% of future TB cases in each migrant cohort. However, because the majority of recent migrants are young adults with relatively low LTBI prevalence, low TB mortality risks and high emigration risks, strategies were unlikely to be cost-effective to the Australia health system unless most of the strategy cost was borne by migrants through pre-migration screening. Furthermore, application of even small quality of life decrements associated with LTBI treatment suggested these strategies would cause more ill-health than they prevent. This thesis draws on both historical and new evidence to improve our understanding of the risk that LTBI poses in the years after infection, and the management of that risk. Additional LTBI screening and treatment strategies cannot currently offer a cost-effective pathway to TB elimination in Australia. However, should any new research come to light, or new tools or treatments for LTBI become available into the future, the model presented in this thesis will be able to reassess this. In their absence, existing TB control strategies in Australia have ensured that TB incidence has remained stable despite decades of immigration from high-incidence countries, and these strategies should be maintained. Additionally, it is important to remember that the origin of the TB burden in Australia predominantly lies in infection that occurs overseas. Therefore, for as long as we continue to live in an interconnected, globalized world, the most effective current pathway to TB elimination, in both high and low-incidence settings, lies in supporting TB control in high-incidence countries. TB researchers in Australia and other low incidence settings should ensure that the research, messages and guidelines they champion address this need.

Human gamma-delta (gd) T cells can respond rapidly with cytotoxic responses to pathogenic infections and malignant cells. Although they are found at low numbers in peripheral blood, these might expand and constitute up to half of the total circulating T cells after infection. Most systemic gd T cells expressed a recombined heterodimeric T-cell receptor (TCR) with genes of the variable (V) g9 and d2 loci, termed Vg9Vd2 T cells. Opposed to conventional ab T cells that recognise processed peptides on major histocompatibility complex (MHC) class I and II and elicit delayed adaptive responses, gd T cells are innate-like T cells that recognise non-peptidic antigens akin to mucosal-associated invariant T (MAIT) or MHC-like lipid-presenting molecule (CD1)-restricted T cells which recognise vitamin-B derivates or lipids, respectively. The most interesting feature of Vg9Vd2 T cells is that they recognise phosphorylated antigens (aka phosphoantigens) which are essential for life including bacteria and particularly abundant in cancer cells. Little is known how gd T cells recognise phosphoantigens, even though, a few putative ligands have been described. Butyrophilins (BTNs) are a family of transmembrane molecules capable of regulating the immune activity of innate-like gd T cells. They dimerise to constitute complexes, some of which may interact with germline-encoding regions of the gd T cell receptors in murine or human species. For instance, mouse butyrophilin-like (BTNL) molecules shape the Vg7+ gd T-cell compartment in the intestine, while human BTNL counterparts selectively activate Vg4+ gd T cells of the colon. Here, tetrameric gdTCR clonotype probes are used for a genetic screen to identify a previously unknown molecular ligand essential for recognition of phosphorylated antigens by gd T cells. This screen identifies BTN2A1 and subsequent experiments elucidate this protein contacts with the Vg9 domain irrespective of the Vd recombined segment and is essential to confer reactivity to phosphoantigens together with BTN3A1. Thus, BTN2A1 and BTN3A1 constitute a functional complex of which we found both intracellular domains are critically important in maintaining an active conformation. Whereas internal BTN3A1 PRY/SPRY (B30.2) motif senses the antigen, the BTN2A1 intracellular domain appears fundamental to retain association of the complex. Mutagenic alanine screens reveal a dual-ligand binding site for the Vg9Vd2 TCR, where conserved residues of the Vg9 domain bind to BTN2A1 and several residues located at the complementarity-determining regions (CDRs) might react to a putative molecule of the BTN3A family. Thus, this work proposes a phosphoantigen-reactive gd T cells recognise a dual-ligand binding complex where BTN2A1 contacts the Vg9 domain and BTN3A1 is a phosphoantigen sensor molecule that plausibly induces the molecular switch necessary to induce immune responses. Lastly, the influence of tumour infiltrated phosphoantigen-reactive gd T cells in renal carcinogenic tumour patient-derived organoids (PDO) samples is examined and their relevance in healing disease assessed in comparison to previous clinical studies. These results are of vital importance to better understand the potential of gd T cells in prospective medical applications.

I would like to take time now to thank the people that significantly contributed to me reaching this far. Herewith, being so grounded in my faith, I would like to start by thanking God, for being the source of my strength and peace. Secondly, I would like to thank my supervisors Paul Cameron and Sharon Lewin. Paul, thank you for your invaluable insight in formulating the research questions. At first, it was really hard for me to see you retire towards the end of my PhD, as I missed the option, I had of just popping in your office to ask you a question. Nevertheless, thank God for technology as we found a way to make it work. Therefore, I’m grateful to you for providing me with the opportunity to call you whenever I needed your help (including the weekends), not many students have that option. Also, thank you for asking me critical questions and pushing me to resolve issues on my own but also helping me when needed. Although it was time-consuming some days, I am grateful as it made me the independent person that I am today. I also appreciate you supporting me whenever I would tell you about a meeting or conference that I think might benefit me or my work. I am very appreciative of my second supervisor Sharon for allowing me to join her lab after leaving the Department of Medicine at the Austin Hospital to join the Department of Microbiology and Immunology at the Peter Doherty Institute. This was a huge move for me, but it has always been my dream to work on HIV-1/AIDS, so thank you for allowing that transition to happen so smoothly. Your insightful feedback pushed me to sharpen my thinking and brought my work to a higher level. Also, thank you for finding a way to collaborate with my former lab at the Austin Hospital and with other experts in this field such as Jake Estes, who have significantly helped with my PhD. As a somewhat shy person, this meant I was out of my comfort zone and that I would have to meet new people or be in a new working environment. Looking back, I can say that I have grown so much and became independent, and as a result, I was able to foster and maintain collaborations. I would also like to thank Sharon for her support when my health wasn’t optimal, this contributed to my recovery and my eagerness in returning to the lab and finishing. Empathy goes a long way! Sharon, I want to thank you for your support and for all the opportunities I was given to further my research. I would like to also thank Vanessa Evans for co-supervising me after Paul’s retirement. I appreciate you travelling to The Austin Hospital to meet me and your input in my research. You have been also so supportive, thanks for also being a listening ear when I needed it. I would also like to thank Thomas Rasmussen for assisting me towards the end with helping me to organize my results and proof-reading my thesis. I have always been amazed at how quickly you would get back to me, and for that, I am truly appreciative of. I am very thankful for our lab manager Ajantha Solomon and our research manager Jasminka Sterjovski (Peter Doherty Institute). Aj, thanks for assisting me and being so supportive when I was working at the Austin, I will miss our little walks. Jas, you have been so supportive and understanding. I will never forget when the Hurricane destroyed 90% of my island back home and I was unable to contact my family for a few weeks. This was one of the scariest moments in my life, but you were there every day checking to see if I was okay and offering me to go for a walk in the park. I will forever be appreciative of your kindness, empathy and support, you are one of a kind! Special thanks to my collaborators at Olivia Newton-John Cancer Centre Research Institute, Austin Hospital, Australia. Marzena Walkiewicz (my lab aunty), thanks for teaching me all there is to know about immunohistochemistry and for all the delicious treats. I would also like to thank Dani for training me on the Vectra imaging platform and Johnathon Cebon and Katherine Woods for being so graceful when I decided to switch to HIV-1 research and welcoming me back for collaboration, it felt as if I have never left! I would also like to thank my collaborators Claire Deleage and Jake Estes (Frederick National Laboratories for Cancer Research, USA), thank you for picking me up every morning and teaching me RNA/DNAscope and imaging. On the same note, I would also like to thank my other collaborators, Gustavo Reyesteren and Perla Del-Rio (Instituto Nacional de Enfermedades Respiratoriras, Mexico), Michael Gonzales, Samuel Thomson (Pathology Department, The Royal Melbourne Hospital, Australia) and Metta Jena and Rejhan Idrizi (Peter MacCallum Cancer Centre, Australia). In addition, I am greatly appreciative of the immunomodulation group (Peter Doherty Institute), who contributed to my research, particularly Judy and former member Renee, who has trained me so quickly on getting PC3 certified and on the in vitro model of Latency. I also owe gratitude to our research assistants, especially Caroline and former research assistant Surekha for always following up on my orders and contacting me if there were any issues. I would also like to thank present and former members of the lab Simin, Hao, Sandy and Jenny. To my parents, O’Neil Richardson and Julienne van der Leeuw-Richardson, thank you for the encouragement, sacrifices made and for permitting me to leave my Island at a young age to pursue my dreams. To my big sister Nyakomi, thanks for always being my hype man and believing in me, and to my other siblings, Glenroy and Abigail, my intelligent nephew Amaziah and friends overseas and the ones I’ve made in Australia (too numerous to mention), thank you for supporting me in every aspect during the completion of my PhD. After completing my masters, I was fortunate to work in my hometown St. Maarten (which has the largest amount of people living with HIV-1/AIDS in the Dutch Caribbean islands) at the HIV-1/AIDS foundation, interviewing patients suffering in secret. This has also contributed to me pursuing a PhD in this field. Therefore, I would like to acknowledge those living with HIV-1, particularly in places where stigma and discrimination are at the highest. I’ve learned that people will forget what you said, people will forget what you did, but people will never forget how you made them feel.' M. Angelou Thank you for all the positivity and support throughout this journey.

Identification of Mycobacterium tuberculosis (Mtb) increasingly involves characterising large sections of genetic material, such as through whole genome sequencing. While some mutations identified through these techniques are well characterised and strongly associated with anti-tuberculous drug resistance, such molecular methods frequently identify mutations with unknown significance or limited understanding of associated functional biological pathways. In this PhD, I have developed computational protein structural tools and mathematical models of TB transmission, that use genomic data to understand the impact of genomic changes and predict the consequences with regards to transmissibility and drug susceptibility of Mtb. Drug resistant mutations often carry both a selective advantage and a fitness cost, which can be reflected by the changes in protein structure and function. I developed a pipeline that captured the molecular consequences of coding mutations on protein stability, dynamics and interactions. Using my pipeline to evaluate the mechanistic consequences of mutations, I applied it to the real-time genomic analysis of a Victorian tuberculosis patient. The analysis led to identification of a novel resistant strain and altered patient treatment – the first reported use of structural information to guide clinical resistance detection. The information was then used to inform a compartmental epidemiological model of Mtb transmission in order to understand the rise of drug resistance in two high TB-incidence setting. Using a adaptive metropolis algorithm, I estimated drug resistance amplification proportions for two first-line anti-tuberculosis drugs, and explored how structural changes may alter the fitness landscape and transmission dynamics. The work highlighted the power of combining genomic, epidemiological and structural information in the fight against tuberculosis, and presents examples of application across the spectrum from laboratory, clinical and programmatic contexts. This work has further laid the foundation to rapidly apply and translate this approach to other infectious and non-infectious diseases.

Malaria in pregnancy is responsible for 10,000 maternal deaths and 100,000 infant deaths every year. Pregnant women, despite lifetime malaria exposure, are more susceptible to Plasmodium falciparum malaria than their non-pregnant counterparts. This is due to the ability of parasites to sequester in the placenta by expressing VAR2CSA, a pregnancy-associated P. falciparum erythrocyte membrane protein 1 (PfEMP1) variant surface antigen that is upregulated during pregnancy. Naturally acquired antibodies (Abs) to VAR2CSA are associated with reduced adhesion of infected erythrocytes to the placenta. The two leading placental malaria vaccine candidates PAMVAC and PRIMVAC both include the VAR2CSA antigen subdomain, Duffy binding-like domain 2 (DBL2). Although Abs to DBL2 may be important in preventing placental malaria, whether these Abs need to trigger downstream effector functions, such as Ab-dependent cellular cytotoxicity and complement activation, or if the genetic characteristics of the Abs modify these functions, is poorly understood. This PhD project aimed to investigate the Ab response to VAR2CSA, exploring the functional and structural properties of these Abs to identify the mechanisms behind protection. Customised multiplex assays were developed, and systems serology approaches were used to identify key VAR2CSA domain targets of Ab immunity in women with placental malaria compared to non-placental malaria-infected women (infected with malaria, but no sequestering in placenta) from Papua New Guinea (PNG). Machine learning techniques demonstrated that IgG Abs (mainly IgG3) and C1q were the main features discriminating between placental and non-placental malaria (Chapter 2). These could be used to predict Ab-driven functional mechanisms behind protection against placental malaria. Therefore, we hypothesised that stronger Ab responses to the DBL domains will mediate more effective downstream effector functions, such as Natural Killer (NK) cell activation via Fc gamma receptor (FcgR) functions (Chapter 3). However, even though Ab-mediated NK cell activation was observed in pregnant women with malaria, no differences were associated with susceptibility to placental malaria. We also explored the importance of allotypes of immunoglobulin subclasses IgG1 and IgG3 in placental malaria infections. IgG1 and IgG3 allotypes are polymorphisms in the constant regions of immunoglobulin heavy and light chains. IgG3 allotypic variations can have structural and functional consequences, such as shorter hinge regions and extended half-lives, which may play a role in malaria susceptibility and protection in placental malaria. IgG allotypes may also have a significant impact on associated Fc-mediated effector functions. We identified a novel IgG3 allotype, that was more prevalent in an isolated population in PNG, which was associated with reduced IgG3 Ab levels in plasma. We explored the impact of this novel variant on FcgR binding, its engagement with FcgR dimers and its capacity to induce phagocytosis in the context of placental malaria infections (Chapter 4). Glycosylation patterns associated with IgG responses to malaria infections during pregnancy are relatively poorly understood but identifying such patterns could help understand the importance of IgG during placental malaria infections and could aid to develop efficient Ab-based vaccines. Here, we showed that pregnancy-associated anti-inflammatory IgG Fc glycans may dampen the Ab-mediated activation of NK cells in pregnant women with malaria infection (Chapter 3). To date, malaria-specific IgG glycosylation has not been systematically investigated, so this PhD project aimed to establish a method to purify antigen-specific Abs (Chapter 5). A detailed understanding of the glycosylation patterns of antigen-specific IgG Abs and their interactions with FcgRs and/or complement complexes may be key to uncovering novel pathways of Ab-mediated responses in placental malaria infections. In summary, this thesis analysed the functional and structural properties of Abs that could help guide the rational development of future Ab-based vaccines and therapeutic Abs to prevent clinical complications, not just placental malaria.

The immune system can recognise and control cancer cells in a process termed cancer immunosurveillance. There is increasing evidence that CD4+ T cells play an important role in melanoma immunosurveillance but considerable debate surrounds the underlying anti-tumoral mechanisms. This project thus sought to unravel the role of CD4+ T cell responses to melanoma using a transplantable orthotopic murine melanoma model in conjunction with newly generated genetically modified B16 melanoma cell lines. Remarkably, adoptive transfer of naive or activated antigen-specific CD4+ T cells was highly protective against the development of melanoma. In addition to a classical “helper” function, CD4+ T cells acted as peripheral anti-tumoral effector cells whereby they migrated into the skin, differentiated into Th1 cells and mediated local suppression of tumor development. Accordingly, we provide evidence that CD4+ T cells can directly kill melanoma cells in vitro through several cytotoxic pathways, including TNF superfamily signalling via TNF and FasL as well as perforin-dependent cell lysis. Finally, we investigated the role of MHC-II expression by melanoma on the antitumoral function of CD4+ T cells. Whilst MHC-II expression by melanoma cells promoted CD4+ T cell infiltration into the primary tumor site it was dispensable for control mediated by CD4+ T cells. This suggested an important role for indirect display of MHC-II-restricted epitopes by antigen-presenting cells within the tumor microenvironment. This was supported by visualization of melanoma-specific CD4+ T cells in the tumor microenvironment using two-photon microscopy, where activated CD4+ T cells appeared to interact with melanoma cells via intermediary cells, presumably professional antigen-presenting cells. Finally, we observed a reduction in metastatic lesions in the tumor-draining lymph node in mice challenged with MHC-II deficient melanoma cells. These data suggest that MHC-II may play context-dependent roles in control of primary tumors and lymph node metastases by CD4+ T cells. In summary, this study demonstrates the important role of CD4+ T cells in melanoma immunosurveillance and provides important insights into underlying antitumoral mechanisms.

Morbidity and mortality rates from seasonal and pandemic influenza virus infections occur disproportionately in high-risk groups, including Indigenous people globally. Although adaptive immunity is essential for combating viral infections, it is largely understudied in Indigenous populations, including Aboriginal Australians. To rationally design a protective vaccine against both seasonal and pandemic influenza viruses for both Indigenous and non-Indigenous people, in-depth knowledge about B cell and T cell responses is needed. Therefore, the aim of this PhD thesis was to identify and characterise CD8+ T cell responses restricted by HLA alleles dominant in Indigenous Australians as well as understand CD4+ T cell, B cell and humoral responses to the inactivated influenza vaccine in Indigenous and non-Indigenous Australians. The data generated in this PhD thesis provide key insights into a rational design of a cross-protective influenza vaccine. The HLA-A*24:02 allele was associated with a higher mortality during the 2009 influenza pandemic and is highly expressed in Indigenous populations globally. In Chapter 3, we identified novel influenza A and B virus (IAV and IBV, respectively) CD8+ T cell targets presented by HLA-A*24:02 and characterised their immunogenicity in HLA-A*24:02 transgenic mice, and HLA-A*24:02+ Indigenous and non-Indigenous individuals. We discovered the first IBV-derived CD8+ T cell epitopes presented by HLA-A*24:02. A total of 6 novel immunogenic IBV epitopes were identified, with one epitope, A24/PB2550-558 being cross-reactive between IBV and the IAV variant, making it especially important for a cross-protective universal vaccine. We defined epitope-specific CD8+ T cells ex vivo in the blood of healthy Indigenous, non-Indigenous and acutely-infected donors. These newly identified epitope-specific CD8+ T cells were readily detected during acute infection, mainly of an effector memory phenotype and expressed activation markers such as PD-1, CD38, HLA-DR and CD71, thus highlighting their involvement during influenza virus infection. Sub-cutaneous vaccination of transgenic mice with three immunogenic IBV peptides resulted in reduced viral titres, reduced pro-inflammatory cytokines MIP-1a, MIP-1b and RANTES in the lung as well as reduced weight loss in comparison to unvaccinated mice. These data highlight the importance of identifying novel CD8+ T cell epitopes for their potential use as a universal influenza vaccine in the context of a highly prevalent HLA allele in Indigenous people globally. Indigenous Australians express a unique HLA class I profile which includes high frequencies of HLA-A*24:02, A*11:01, A*34:01, B*13:01 and B*15:21. Except for HLA-A*24:02 and HLA-A*11:01, no influenza virus-derived epitopes have been identified for the Indigenous-associated HLAs. In Chapter 4, we generated single HLA-expressing cell lines to use as antigen-presenting cells for identifying and characterising CD8+ T cell responses towards the Indigenous-associated HLAs. We optimised an influenza virus infection assay, instead of using peptide pool cultures, to stimulate and expand rare influenza-specific CD8+ T cells, and thus allow sufficient numbers for the sequential dissection of single epitope candidates. As such, we identified the first IAV and IBV epitopes presented by HLA-B*13:01. While responses to IAV were dominated by the highly variable B13/NP404-412 epitope, IBV-responses were directed towards a variety of epitopes, including the dominant and highly conserved B13/HA371-379 and B13/HA427-435 peptides. We sequenced IAV- and IBV-specific CD8+ T cell receptors and identified unique TCR signatures between the different epitopes. While B13/NP404-412-specific CD8+ T cells showed a limited bias for the expression of TRBV19 with a public TCR, B13/HA371-379-specific CD8+ T cells almost uniquely expressed TRAV3 paired with a variety of different TCRb-chains. The developed tools proved to be an efficient method to identify and characterise IAV and IBV epitopes presented by Indigenous-associated HLAs to further develop a universal influenza vaccine that can protect Indigenous Australians. Current influenza inactivated vaccine strategies delivered intramuscularly induce mainly B cell and antibody responses. Responses to the influenza vaccine are studied extensively but not fully understood in high-risk populations. Despite the strong recommendations for annual influenza vaccination, Indigenous Australians are completely underrepresented in studies that analyse vaccine responses. In Chapter 5, we defined cellular and humoral responses to the inactivated influenza vaccine in Indigenous Australians vaccinated between 2016 and 2018, and compared these responses to non-Indigenous Australians. We identified robust antibody responses to the vaccine that were comparable to non-Indigenous donors and were cross-reactive with viruses circulating more than 10 years ago. Antibody responses on day 23+ post vaccination correlated with acute activation of circulating T follicular helper cells type 1 (cTFH1) in line with previous studies. System serology of selected Indigenous and non-Indigenous donors revealed significant reduction of the contribution of IgG3 to influenza-specific antibodies in Indigenous Australians, which correlated with a higher frequency of the G3m21 allotype. The generated data are important to support vaccine recommendations for Indigenous Australians, but also highlight the need to improve influenza vaccinations by harnessing the protective capacity of CD8+ T cells in future vaccine designs. Overall, this PhD thesis provides highly important knowledge of T and B cell immunity to IAV and IBV in Indigenous Australians. The findings of this PhD thesis provide key insights into the development of a universal influenza vaccine that also protects Indigenous Australians, one of the high-risk groups of developing severe influenza disease.

Vaccination is the best available means to reduce the burden of seasonal influenza. However, current influenza vaccines need to be updated frequently to keep up with evolution among circulating viruses. Antigenic evolution, otherwise termed drift, is most rapid among A/H3N2 viruses, and the A/H3N2 component of vaccines is frequently updated. Despite this, influenza vaccine effectiveness against the A/H3N2 subtype has been poor in recent years, especially among previously vaccinated individuals. Protection induced by inactivated influenza vaccines is largely mediated by B cells and antibodies reactive against the head of the hemagglutinin (HA) protein, with help from T follicular helper cells. The cellular and molecular mechanisms that underlie the attenuating effects of prior vaccination and existing immunity are largely undefined. It has been suggested that existing antibodies clear or mask antigen, or that memory B cells induced by prior exposures competitively dominate responses so that B cells and antibodies become focused on epitopes that are shared between prior and prevailing vaccine strains. The aim of the work presented in this PhD thesis was to examine the impact of pre-existing immune responses induced by prior infection with different A/H3N2 strains on influenza vaccine immunogenicity. In depth antibody as well as B cell assessments were performed to understand the impact of existing antibodies and memory B cells following vaccination and provide insights into the design of new vaccine strategies. As a lead up to the ex vivo analysis of B cells from vaccinees, we first sought to understand how human naive versus memory B cells differentiate in vitro. Experiments were conducted in Chapter 3 to compare the stimuli required for their differentiation into plasmablasts, and subsequently understand how they change phenotypically once stimulated. Specifically, sorted human naive and memory B cells from healthy individuals were stimulated in vitro to induce differentiation into plasmablasts. Data obtained in this PhD thesis showed that stimulation with the Toll-like receptor (TLR) 7/8 agonist R848 in the presence of monocytes induced the highest activation of both naive and memory B cells. Conversely, stimulation with the TLR9 agonist CpG or with R848 in the absence of monocytes induced little to no differentiation of naive B cells but were able to stimulate memory B. cell differentiation. Despite robust differentiation into antibody secreting plasmablasts, naive-derived B cells remained phenotypically distinct from memory-derived B cells up to day 6 after in vitro activation, with differential expression of CD27, CD38 and CD20. This work resulted in a first-author publication in Clin Transl Immunol, 2019. The focus of Chapters 4 and 5 was to understand how prior influenza virus infection affects antibody and B cell responses to influenza vaccination. To address this question, vaccine responses were investigated in a unique influenza vaccine-naive cohort in Viet Nam, that had been monitored for both clinical and asymptomatic influenza virus infection for more than 9 years. In 2016, twenty-eight participants without documented A/H3N2 virus infection (since 2007) and 72 participants who had been infected with A/H3N2 viruses, belonging to a range of genetic clades, received an inactivated trivalent influenza vaccine containing an A/Hong Kong/4801/2014-like (H3N2) antigen. This work investigated whether influenza vaccination induced naive B cell responses specific for new epitopes or largely recalled B cells specific for conserved epitopes, common to the vaccine A/H3N2 component and prior infecting strains. Hemagglutination inhibition antibody titres were measured in pre- and serial post-vaccination sera against 40 A/H3N2 viruses spanning 1968-2018 to understand how the titre and cross-reactivity of antibodies against the HA head evolve. B cells were assessed by flow cytometry using a panel of phenotypic markers in addition to recombinant HA probes representing the vaccine and recently infecting strains (A/Perth/16/2009, A/Victoria/361/2011 and A/Switzerland/9715293/2013). Participants who had at least one pre-vaccination A/H3N2 virus infection had on average 2 to 3-fold higher vaccine-specific antibody titres, steeper titre rises in the weeks following vaccination (mean peak on day 14), and less titre decay by days 21 and 280 compared to participants without prior infection. Moreover, participants with prior infection exhibited greater and better-maintained titre rises against viruses that circulated a year after vaccination, indicating that prior infection extends the strain coverage of antibodies induced by vaccination. Notably, A/H3N2 viruses that circulated 275-340 days after vaccination caused illness in only 1.4% of participants with infection prior to vaccination and in 14% of participants without prior infection. This suggests that vaccine effectiveness can be enhanced by pre-existing immunity. However, it was also clear that the range of strains against which antibodies were induced was dictated by the strain with which participants were previously infected, indicating that vaccination may simply recall rather than update antibody-mediated immunity. HA-probe reactive B cell frequencies and activation status increased substantially after vaccination. The greatest increases in HA probe-reactive B cells were detected among participants who had recent prior infection, with the majority of B cells exhibiting cross-reactivity with prior strains. A modest but significant increase in the frequency of B cells that reacted with the HA of the vaccine strain, but not of past strains, could be detected in participants who lacked prior infection. The phenotype of vaccine HA single-positive B cells, including increased IgM expression, indicated that they may have been naive-derived B cells. Vaccination induced B cells that preferentially reacted with the HA of A/Perth/16/2009 and/or A/Victoria/361/2011 viruses, but not A/Switzerland/9715293/2013 viruses, among participants who had prior A/Perth/16/2009-like virus infection. However, B cells induced by vaccination in participants who had prior A/Switzerland/9715293/2013-like virus infection were equally cross-reactive with HA of all tested viruses. These results support the inference that immune responses to standard inactivated influenza vaccines are dominated and shaped by recalled memory B cells with limited activation of naive B cells to update immunity. Overall, this PhD thesis investigated how pre-existing immunity induced by documented influenza virus infection affected the humoral response to seasonal influenza vaccines in healthy adults. This work provides new insights into the capacity of influenza vaccines to stimulate naive B cells, which may be limited due to memory B cell dominance and to a lack of sufficient stimulation to activate naive B cells. This knowledge could be used to design new vaccine strategies and improve influenza vaccine-induced protection.

Seasonal influenza viruses circulate globally and cause recurrent disease in humans. Worldwide, annual epidemics are estimated to cause 1 billion infections, with 3 to 5 million cases of severe illness and 290,000 to 650,000 deaths. Influenza viruses undergo rapid antigenic evolution allowing mutant viruses to escape from host immune responses acquired to parental virus strains. Current seasonal influenza vaccines are effective when vaccine strains are matched with circulating strains. However, there is little to no cross-protection against antigenic variants, emerging pandemic or zoonotic outbreak strains. There is therefore tremendous interest in the development of novel universal vaccines which induce potent, broad and durable antibody responses against most or all influenza viruses. T follicular helper cells are crucial for the generation of high affinity antibodies and the maintenance of B cell memory. But relatively little is known about Tfh in important animal models of influenza. Insights gained from the study of Tfh cell responses will facilitate the design of next generation vaccines against influenza. In this thesis, we first developed an activation-induced marker assay for the identification of antigen specific Tfh cells in mice after influenza virus infection and hemagglutinin protein immunization. We showed that the AIM assay was robust and sensitive for the detection of murine Ag specific Tfh cells by quantifying the upregulation of surface CD154 or CD25 OX40 following either HA peptide pool or whole HA protein stimulation for 18 hours. This murine AIM assay makes it feasible to delineate Ag specific Tfh cells in mice without the need for transgenic mice or MHC II tetramers restricted to specific epitopes. Importantly, Ag specific Tfh cells can be sorted for TCR sequencing or adoptive transfer since AIM assay is a live cell assay. Ferrets are a well established animal model for influenza research and are widely used to investigate the pathogenesis and transmission of influenza viruses and preclinically evaluate the efficacy of influenza vaccines. However, little is known about ferret Tfh cells due to the lack of ferret reactive immunological reagents. To enable the study of ferret Tfh cells, we screened commercial markers of Tfh cells, antiBCL6, CXCR5 and PD1 antibodies, and found two anti-BCL6 antibodies had cross reactivity with lymph node cells from influenza infected ferrets. We also developed two murine monoclonal antibodies against ferret CXCR5 and PD1 using a single B cell PCR based method. We were able to clearly identify Tfh cells in LNs from influenza infected ferrets using these antibodies. The development of ferret Tfh marker antibodies and the identification of ferret Tfh cells will facilitate the assessment of vaccine induced Tfh responses in the ferret model and the design of novel vaccines against influenza infection. HA stem is an attractive target for the development of universal influenza vaccines due to its relatively conserved feature. However, HA stem is poorly immunogenic when administered alone in a soluble form. Immunogen multimerization can enhance the immunogenicity of poor immunogens even in the absence of the help of T cells, which serves as an alternative pathway to improve the immunogenicity of stem without the dependence on Tfh responses. We showed that chemically coupling a peptide derived from the head domain of PR8 HA, P35, with the weakly immunogenic HA stem protein caused aggregation of the HA stem which significantly enhanced stem specific B cell responses independent of Tfh cell help in mice. P35 conjugation represents a new pathway to boost stem specific antibody responses without introducing exotic carrier proteins which will elicit anti carrier responses. Collectively, we investigated Tfh responses to influenza virus infection and immunization in mice and ferrets and explored the effects of immunogen multimerization on humoral immunity in the context limiting Tfh responses to HA stem. An increased understanding of Tfh dependent and independent mechanisms to enhance humoral immune responses will assist developing novel vaccines to prevent the infection of influenza and other viruses.