HIV-1 capsid (CA) plays a vital role in the early stages of HIV-1 infection. The CA lattice surrounding the viral core is predominantly assembled from CA hexamers and stabilised by intra-hexamer and inter-hexamer interactions. Optimal stability of the lattice is known to be critical for efficient infection; however, a comprehensive screen of the effects of stabilising all lattice interfaces has not been performed. Disulphide cross-linking of residues across lattice interfaces has been used in vitro to stabilise CA assemblies. In this study, putatively stabilising cysteine CA mutations were designed at each interface of the CA lattice and their effects on early life cycle events, including reverse transcription and nuclear entry, assessed. The introduction of cysteine mutations at intra-hexamer (both NTD-NTD and NTD-CTD) and inter-hexamer (dimeric CTD-CTD only) lattice interfaces resulted in cross-linking and hyperstable viral cores in infected cells. These cores were minimally infectious and encountered sequential blocks to infectivity at reverse transcription, nuclear entry and post-nuclear entry. The infectivity defect of hyper-stable core mutant, A14C/E45C, was partially compensated – without an observable decrease in stability – by addition of mutations reported to perturb interactions with CPSF6. In contrast, Nup153 and CypA mutations were unable to compensate the infectivity defect suggesting that this was a CPSF6-specific effect. Proximal ligation assays were performed to visualise and quantify interactions between CA and host factors, indicating that hyper-stable cores encountered a block to nuclear entry in G1/S arrested cells. Overall, the results of this study suggest that mutations at different lattice interfaces can result in global changes to the intrinsic stability of the viral core and results in fitness defects at multiple stages of the HIV-1 early life cycle.

We hypothesise that HIV-1 DNA synthesis occurs inside intact viral capsid (CA) cores. We propose that dNTPs are transported into the CA via an electrostatic channel, formed by six positively charged arginines in the centre of CA hexamers. Here, we consider whether reverse transcriptase inhibitor (RTI) sensitivity is altered when the nature of the channel is changed, either by Gag mutation or exchange with Gag from a non-pandemic HIV isolate with a different structure. There are two classes of RTIs: nucleoside/nucleotide based (NRTI) and non-nucleoside inhibitors (NNRTI). We hypothesised that negatively charged NRTIs would recruit to CA hexamers, to be transported into cores. However, NNRTIs are uncharged, yet potently inhibit DNA synthesis, suggesting that NNRTIs enter cores by diffusion or inhibit after uncoating. We tested HIV-1 vector sensitivity to RTIs, either bearing lab adapted M-group, transmitted founder, O-group or mutant Gag sequences. Viral inhibition was measured by comparing IC50 and IC90 values in a range of cell lines. Our data shows that some differences in Gag demonstrate a cell type-dependent effect on viral sensitivity to RTIs. We also tested the stage of RTI inhibition, measuring early and late-reverse transcription (RT) products of HIV-1 (M) and HIV-1 (O) virus in the presence of inhibitors. Our data show that both HIV-1 (M) and (O) vectors are inhibited after 2nd DNA strand transfer. We determined that a small number of vDNA strands are required to infect a U87 cell, which increases in the presence of RTIs or on R18G mutation. We conclude that differences in Gag have some small cell type-dependent effects on RTI sensitivity. We hypothesise this may be due to differences in the timing of CA uncoating between cell types, supported by our finding that all RTIs tested inhibit RT predominantly after 2nd strand transfer.

Tuberculosis (TB) poses a significant challenge for global health. Developing countries most burdened by disease have insufficient laboratory resources and limited capacity to conduct clinical trials. This thesis aims to investigate microbiological methods currently employed in TB programmes to compare their value in diagnosing TB and monitoring treatment using data from smear-positive patients enrolled in the REMoxTB study. As TB culture remains out-of-reach in many settings, diagnosis continues to depend on sputum smear microscopy. Early morning samples are considered better than spot samples. However, the data presented show that spot samples have a higher positive yield and greater sensitivity for solid and liquid culture. This evidence does not support guidelines requiring early morning samples which inconvenience patients and complicates trial enrolment. The data also show a reducing correlation between smear microscopy and culture on solid and liquid media as treatment progresses. These findings question the use of smear microscopy as a proxy for culture during treatment which may prompt inappropriate treatment extensions or retreatments. Comparing solid and liquid culture, the analyses show that liquid culture is faster and more sensitive. The clinical significance of this increased sensitivity throughout treatment is uncertain, and some samples negative in liquid culture remain positive on solid culture. The relationship between solid and liquid culture changes during treatment suggesting they differentially support the metabolic requirements of changing mycobacterial populations. The value of measures of pre-treatment mycobacterial load are investigated and show they are poorly predictive of microbiological responses during treatment in either culture media. Better indicators of treatment response are required which reflect mycobacterial population dynamics. Increasing availability of TB culture would of great benefit. The data show that the incubation times for liquid cultures, particularly for diagnostic samples, may be significantly reduced without loss of sensitivity, which could increase laboratory capacity and remove barriers to implementation. Future work will investigate whether these findings are generalisable to smear-negative patients and assess their value in predicting long-term treatment outcomes.

Activated PI3Kδ syndrome (APDS) is a recently described primary immune deficiency characterised by gain-of-function (GoF) mutations in the PIK3CD gene which encodes the leukocyte-restricted p110δ catalytic subunit of phosphoinositide 3-kinase enzyme (PI3K) known to be important in cell survival, growth and migration. So far, defects of T and B-cell function have been reported but other immune cell types have not yet been well studied. Although p110δ is known to have a role in myeloid cells, the impact of APDS on innate immunity is currently unknown. Here, using both human APDS samples and a murine model of APDS, we demonstrate that the E1021K GoF mutation in p110δ is associated with enhanced PI3K signaling basally as well as upon TLR4 ligand lipopolysaccharide (LPS) stimulation in dendritic cells (DC). Surprisingly, no effect of the mutation was observed on cytoskeletal function. However, APDS DC displayed altered LPS signaling, resulting in impaired IL- 12 and IFN-β secretion. In a murine model of APDS, reduced IFN-β secretion by DC in turn led to diminished autocrine and paracrine IFN-β signaling, which culminated in defective production of anti-microbial nitric oxide. We hypothesise that this may contribute to increased susceptibility of APDS patients to bacterial infections, warranting further investigation.

Vesicular stomatitis virus Indiana strain is the prototype envelope glycoprotein for the genus Vesiculovirus. While the wild-type virus, VSVind, has developed into a potent and versatile oncolytic virotherapy and vaccine vector delivery platform, the G protein (VSVind.G) is ubiquitously used to pseudotype lentiviral vectors (LVs) for experimental and clinical applications. Recently, G proteins derived from other vesiculoviruses (VesG), for example, Cocal, Piry, and Chandipura viruses have been proposed as alternative LV envelopes with possible advantages compared to VSVind.G. However, vesiculovirus research has not developed extensively, and there is a gap in knowledge regarding the antigenic and immunogenic characteristic of VesG. In this work, I investigated two anti-VSVind.G monoclonal antibodies for their ability to cross-react with other VesG, identified the epitopes they recognise, and explored the mechanisms behind their neutralisation activity. Furthermore, these G proteins were characterised for their sensitivity to inactivation by fresh mammalian sera. Using some mix-and-match constructs, I identified that the hypervariable PH domain of VSVind.G confers sensitivity to otherwise serum resistant Cocal G. I further examined VesG regarding their immunogenicity, explored the humoral immune response triggered by systemic administration of LVs and investigated the inhibitory effects of the induced anti-G neutralising response on subsequent LV administrations. However, this could be alleviated using a heterogeneous panel of envelopes sequentially. Taken together, this work will broaden the use of VesG pseudotyped lentiviral vectors in clinical gene therapy by providing the proof-of-concept to circumvent anti-envelope immunity where repeated systemic vector administration is necessary and the opportunity to modify the VesG and improve G protein-containing advanced therapy medicinal products and vaccine vectors.

Streptococcus pneumoniae serotype 1 displays some unusual epidemiological, clinical and microbiological characteristics. In Africa, serotype 1 is the leading serotype causing invasive pneumococcal diseases, including, pneumonia, meningitis and sepsis. Unusually, serotype 1 is rarely detected during routine nasopharyngeal swabbing, even in areas of high disease burden. S. pneumoniae are natural commensals of the nasopharynx and colonisation of this area has been described as a pre-requisite for invasive disease, hence little is understood about why serotype 1 carriage rates are so low, whilst burden of disease is so high. This body of work sought to identify key virulence factors associated with serotype 1 disease pathogenesis and the effect these have on nasopharyngeal colonisation and progression from carriage to invasive disease. Murine models of nasopharyngeal carriage, pneumonia and sepsis were used to study serotype 1 pathogenesis and identified the bacterial toxin pneumolysin, and autolysin as key virulence factors associated with disease pathogenesis. In addition, the influence of serotype 1 infection on the host immune responses in murine models of pneumococcal infection was also addressed. Current pneumococcal vaccines (PCV13) include serotype 1 however, the impact of this vaccination in reducing serotype 1 disease burden is still unknown. Findings here suggest that pneumolysin, not bacterial capsule should be a target for strong consideration in future vaccine design. New therapeutics should also be designed to target pneumolysin as it is in the main driver of pathogenesis in the context of pneumococcal disease, particularly in serotype 1 infection.

The phosphoinositide 3-kinase (PI3K) signalling pathway is activated in macrophages in response to many bacterial pathogens, triggering phagocytic uptake mechanisms and phosphorylation-associated activation of the serine/threonine kinase AKT. Enteropathogenic E. coli (EPEC) inhibits both PI3K mediated phagocytosis and AKT phosphorylation; dependent on a type 3-secretion system (T3SS) critical for delivering up to 24 known effector proteins into target cells. The efficient translocation of most EPEC effectors is dependent on the T3SS effector chaperone CesT. Although the effectors and mechanisms for inhibiting phagocytosis are well described, little is known how EPEC inhibits AKT phosphorylation. AKT activation is a multi-step process involving its recruitment to the cell membrane and phosphorylation of Thr308 by PDK1 and Ser473 by mTORC2. This activation process is supressed by inositol (such as PTEN) and protein (such as PP1 & PP2A) phosphatases. Altered AKT signalling is associated with many cancers, diabetes, cardiovascular and infectious disease, thus identifying how EPEC inhibits AKT activity could provide insight into its complex regulatory process and/or new therapeutic strategies. Screening of bacterial strains, lacking or expressing subsets of EPEC effectors, by western blot analysis suggests that the inhibitory mechanism depends on the CesT chaperone but not the function of the 21 most studied effectors. The EPEC inhibitory mechanism was investigated through the development of a two-wave infection model, examining for T3SS dependent changes in AKT phosphorylation (Thr308 & Ser473), membrane localisation and activity of AKT associated signalling proteins (PDK1 & PTEN). This strategy revealed inhibition of AKT phosphorylation to be stable (up to 3 h) and linked to increased activity of serine/threonine protein phosphatase(s). This finding was supported by phosphatase inhibitor studies, suggesting the involvement of a host activated or bacterial delivered protein phosphatase. Thus, this study provides new insights into the requirement of the EPEC effector repertoire and suggests a novel mechanism by which EPEC inhibits AKT signalling.

This thesis has been prepared in a paper based format and includes an empirical paper, a systematic review and critical reflection. As a whole, the thesis focuses on the importance of sleep in Chronic Fatigue Syndrome (CFS). The systematic review is prepared for submission to 'Sleep Medicine Reviews'. The empirical paper is prepared for submission to the 'Journal of Consulting and Clinical Psychology'. Paper 1 is a systematic review and narrative synthesis of the current evidence for the effectiveness of Cognitive Behavioural Therapy (CBT) and Graded Exercise Therapy (GET) on sleep in CFS. Eight studies were found and their methodological quality varied. To understand heterogeneity in findings, information regarding intervention delivery, including the presence of sleep management components, methodology and sleep outcome measures was extracted and synthesised. We conclude that GET can improve sleep, when delivered by experienced therapists in outpatient settings. The evidence for CBT on sleep is limited, moreover, at present we know little about the effectiveness of adding sleep management components to interventions. We suggest that sleep outcomes used previously have not been sufficiently comprehensive and sensitive to measure change in sleep difficulties experienced in CFS. Implications for further research are discussed. Paper 2 presents an empirical study examining the relationship between sleep and fatigue in CFS using a daily diary approach. Sleep was measured objectively using actigraphy, and subjectively, using sleep diaries, in order to test which parameters better predict next-day fatigue. We also examined whether negative mood could mediate these relationships and whether subjective sleep variables were predicted by pre-sleep arousal. Using multilevel modelling, we found that subjective sleep, and not objective sleep, predicted next-day fatigue and these relationships were partially mediated by negative mood on waking. Pre-sleep cognitive and somatic arousal predicted subjective sleep variables including sleep efficiency and quality. Based on these findings, we suggest that interventions targeting subjective sleep, such as CBT for insomnia, may be useful in improving experiences of fatigue in CFS. Paper 3 is a critical reflection on the systematic review and empirical research, and on the process as a whole. Strengths and weakness of Paper 1 and Paper 2 are discussed, in addition to consideration of their contribution to wider research, and clinical practice.

HLA-G is a non-classical human MHC I molecule which is abundantly expressed in trophoblast during pregnancy. This molecule is also expressed by several leucocyte subsets but on a low percentage of cells. It can also be secreted in soluble form. The protein is upregulated in several conditions like viral infections, autoimmune disorders and tumours. Several reports have indicated the expression of HLA-G is associated with allograft acceptance and survival. This study was aimed at investigating the expression of HLA-G and its ligand, KIR2DL4, in association with other activation markers, particularly CD103, BAFF and BAFF-R, in relation to CMV antigen stimulation in healthy subjects. Also, HLA-G and KIR2DL4 expression and their association with genotype were studied in renal transplant patients before and after transplantation. HLA-G was expressed by a significantly higher proportion of CD19+ B cells and CD56+ T cells following stimulation of peripheral blood mononuclear cells (PBMCs) from CMV+ healthy subjects with CMV antigen. Also, sHLA-G levels were significantly elevated in PBMC supernatants from CMV+ subjects stimulated with CMV antigen. The HLA-G ligand KIR2DL4 was significantly upregulated in IL-2 culture on CD56+ T cells and significantly downregulated in CD56bright cells. Induction of PBMCs with CMV antigens significantly increased the proportion of CD8+ T cells, NK cells and CD56+ cells T cells expressing this ligand. Also, proportions of HLA-G+CD103+ T cells were significantly increased following CMV stimulation. Stimulation of PBMCs with CMV antigen demonstrated significantly increased proportions of BAFF-R+ CD56+ cells and significantly decreased proportions of BAFF-R+ B cells. HLA-G was expressed on a significantly greater proportion of BAFF+ and BAFF-R+ cells following CMV stimulation. In renal transplant patients, HLA-G was expressed on a significantly higher proportion of CD4+ cells and monocytes as well as B cells following transplantation. However, KIR2DL4 was significantly upregulated only on CD56dim cells following transplantation. Finally, culture of PBMCs with the standard immunosuppressive drugs used in transplant patients did not generally inhibit or augment the effects of CMV on HLA-G, CD103, BAFF and BAFF-R expression by PBMCs from healthy subjects, apart from CD4+ cells and CD56+ cells of which significantly greater proportions expressed HLA-G in response to tacrolimus (Prograf) in combination with CMV particles. Also, CD103 was expressed by significantly higher proportions of these cells in the same treatment combination. In conclusion, the results are consistent with CMV enhancing HLA-G expression on several cell types which may play an important role to allow immune escape during virus infection. In renal transplant patients, proportions of HLA-G+ cells increased significantly in several cell subpopulations after transplantation. Immunosuppressive drug treatment may have contributed to this.

Hypothalamic obesity (HO) is a relatively rare cause of obesity within the population as a whole, but studies of patients with hypothalamic damage show there is a significant prevalence of obesity within the patient group. Additionally, the obesity is difficult to prevent and manage and increases morbidity and mortality in an already at risk patient group. The work of this thesis had two main objectives. Firstly, to examine the prevalence of obesity and associated morbidities in patients with acquired, structural hypothalamic damage with a descriptive cohort study (n=110). A separate descriptive study quantified obesity and metabolic risk factors in young patients with pituitary tumours, with or without hypothalamic damage (n=41), as they were also identified as a possible at risk group during clinical observations. The second objective was to investigate the underlying pathophysiology of HO using various complementary techniques to study patients with hypothalamic damage who remained weight-stable (HWS), patients with HO and age- and BMI-matched controls. These cross-sectional case-control studies used functional magnetic resonance imaging (fMRI; 9 HO, 7 HWS, 20 controls), the universal eating monitor (UEM; 6 HO, 6 HWS, 9 obese controls [OC], 10 non-obese controls [NOC]), Three-Factor Eating Questionnaire (TFEQ; 8 HO, 6 HWS, 9 OC, 11 NOC) and three-day food diaries (6 HO, 7 HWS, 8 OC, 11 NOC) to assess eating behaviour. The first descriptive study included 110 adults with tumours causing hypothalamic damage attending a specialist neuroendocrine clinic. There was a significant prevalence of weight gain and obesity; 81.8% were overweight/heavier, 56.4% obese and 13.6% morbidly obese, despite proactive assessment and treatment during routine clinic visits. Hypertension (30.9%), dyslipidaemia (54.5%), type 2 diabetes mellitus (T2DM) (14.5%) and cardiovascular disease (9.1%) were also prevalent. In 41 patients with childhood/adolescent-onset pituitary adenomas there was also a relatively high prevalence of obesity (39.0%) and cardiovascular risk factors (2 receiving antihypertensive medications, 2 with T2DM and 4 with treated dyslipidaemia), despite the majority of tumours being microadenomas (i.e. too small to cause hypothalamic damage and indicating the need for long-term follow-up of these patients. The first study into the pathophysiology of HO involved the use of functional MRI in 9 patients with HO, 7 HWS and 20 age- and BMI-matched controls. Participants underwent fMRI scans in a fasted state, as well as one hour and three hours following a fixed-load breakfast (25% of their calculated basal metabolic rate [BMR]). At each scan session participants viewed alternating blocks of photographs of high- or low-calorie food, with non-food photographs also viewed to use as a baseline comparison, to allow purely visual activation to be subtracted from any BOLD signal differences which occurred. Whole-brain statistical analysis revealed significantly lower BOLD signal in HWS participants compared to HO (and to controls) in the food motivation and reward-related brain regions of the posterior insula and lingual gyrus (p=0.001) when viewing high-calorie food photographs (compared to non-food photographs). These differences in reward-related brain regions may be implicated in the development of HO/the ability to remain weight-stable despite hypothalamic damage. Eating behaviour studies were undertaken on a separate study day where participants were asked to eat an unlimited pasta meal until adequately full, while seated at the UEM. This allowed monitoring of total intake, eating duration, eating rate and intra-meal on-screen ratings of hunger, fullness and meal pleasantness using visual analogue scales. Additionally Three Factor Eating Questionnaires (TFEQs) and three-day MRC-Human Nutrition Research diaries were completed at home to assess more long-term real-world eating habits. None of the eating behaviour studies identified significant statistical differences between HO and HWS, but this may have been due to lack of statistical power. There was however an unusual pattern of eating rate in those with HO on visual ascription. This involved an initial tendency towards a higher eating rate, followed by a reduction in rate, with a further increase towards the end of the meal. Further investigation of this pattern with larger numbers of participants would be important to determine whether it is a significant finding or merely an anomaly due to the small group size. Interestingly controls ate significantly more at the UEM (even when intake was adjusted according to fat-free mass or as a proportion of the estimated BMR) and for significantly longer than participants with hypothalamic damage, who reported lower hunger at the start of the meal, but there was no difference between obese and non-obese participants. In keeping with previously published research disinhibition scores measured using the TFEQ were higher in participants with simple obesity [1], with no evidence of increased disinhibition in HO participants [2], although the small numbers studied should be noted. Finally, during both of the study days blood sampling was undertaken to look for biochemical/hormonal variances between the groups. Fasting and area under the curve (AUC) active ghrelin concentrations were significantly higher in controls than in patients, in keeping with some (but not all) previous studies [2-4]. Consistent with previously published research leptin concentrations were significantly higher in obese compared to non-obese participants The small size of this study was a significant limitation and limits the generalisability of the findings, particularly in the eating behaviour studies. This was due in part to the rarity of the condition - the occurrence of acquired, structural hypothalamic damage is relatively rare and the number of individuals with hypothalamic damage who remain weight stable is small, leading to particular difficulty in recruiting patients for the HWS group. Whilst there were some interesting findings further larger studies should enable greater clarification and would allow correlation between clinical and biochemical findings. Further studies in larger cohorts could explore the unusual eating pattern seen in those with HO when studied using the UEM and the apparent lack of disinhibition seen in this group. Although this research provides some preliminary novel evidence to support differences in BOLD signal in reward-related regions of the brain as a possible driver of HO, the mechanisms through which these differences exert their effects to contribute to the development of HO require further elucidation and further study with larger numbers of patients is needed.