Introduction and aims: Helicobacter pylori establishes a chronic infection in the stomach of almost half the world's population. A small proportion of patients go on to develop peptic ulcer disease and in some cases gastric cancer. Numerous factors influence the outcome of infection, including host, bacterial and environmental factors. The impact of bacterial colonisation density on the outcome of disease is poorly understood. The aim of this study was to investigate the determinants of bacterial density and the implications in disease. Methods: Gastric biopsies were collected from 50 H. pylori infected patients attending the Queen's Medical Centre, Nottingham for a routine upper GI endoscopy. Patient disease status was determined as having gastritis, duodenal ulcer disease (DUD), or gastric ulcer disease (GUD). Histological analysis for inflammation, activity, H. pylori density, atrophy and intestinal metaplasia were assessed according to the Sydney scoring system. Bacterial strains were cultured from gastric biopsies. Isolated bacterial DNA was used for genotyping strain virulence for CagA, VacA and DupA status. In order to determine accurate levels of colonisation in gastric tissue samples, a quantitative real-time PCR (qPCR) assay was developed based on the quantification of H. pylori 16S rRNA copy number, relative to the number of human GAPDH copy numbers. Quantification of IL8, ILIO, TGFß, FOXP3, h8, Dl and LL37 mRNA levels in gastric tissue were assessed by real-time RTPCR. The impact of host factors on bacterial density was assessed in vitro using H. pylori infected or uninfected AGS gastric epithelial cells, treated with recombinant cytokine. Colony forming units (CFUs) were enumerated in culture suspensions. The in vitro model was also used to examine the impact of cagA, cage, and vacA isogenic deletion mutant strains on bacterial density. Also, h, 8D1 expression levels were measured in culture suspensions, and IL-8 secretion levels in were assessed by ELISA. Results and conclusions: The pattern of colonisation in the gastric mucosa was found to be associated with the disease profile that develops, with an antral-predominant colonisation was associated with DUD, whereas corpus-predominant colonisation was associated with GUD. Bacterial factors that may drive the pattern of bacterial density in the stomach were then investigated, and strains expressing the dupA virulence factor were coincident with antral-predominant colonisation in the gastric mucosa. cagA and vacA were not found to be associated with bacterial density in vivo or in vitro. To assess the relationship between colonisation density and the host response, firstly IL8 mRNA expression levels were measured in gastric biopsies. An inverse relationship between bacterial load and expression of this pro-inflammatory cytokine was observed, but only at high IL8 expression levels and with higher H. pylori density. The inflammatory cytokines IL-10, TNFa, IL-17 and IL-22 did not impact CFU counts in vitro. No association was observed between bacterial load and expression levels of the immunosuppressive genes ILIO, TGFBI and FOXP3. However, increased CFU counts were found on treatment of H. pylori infected AGS cells with recombinant IL-10 and TGFß in vitro. Also, reduced expression of these immunosuppressive genes was associated with the presence of atrophy and intestinal metaplasia. In addition, ILIO and FOXP3 expression levels were negatively associated with mononuclear cell (MNC) infiltration. Reduced expression levels of the antimicrobial peptide (AMP) defensin hß, Dl were found with H. pylori infection in vivo and in vitro, and a negative association between hßDl expression and bacterial load was observed in vivo. A small increase in hßDl expression was observed in AGS cells infected with the cagPAI defective mutant strain in comparison to the wild type strain, suggesting a possible role for the CagPAI in modulating hßDl expression. In contrast, expression levels of the antimicrobial peptide LL37 were upregulated in the H. Mori infected gastric mucosa. In conclusion, the localisation of H. pylori in the gastric mucosa is associated with the disease profile that develops in the stomach. The patterns of colonisation observed may be influenced by the dupA virulence determinant. Host immune responses, were not associated with colonisation density. However, increased Treg response may be associated with protecting the host from the deleterious effects of inflammation. The direct modulation of bacterial load may be regulated by AMPs. In the H. pylori infected mucosa, reduced hßDl expression was observed which may assist persistent colonisation.

Bdellovibrio bacteriovorus is a ubiquitous Gram-negative 8-proteobacterium that is predatory on other Gram negative bacteria. It's prey includes many pathogenic bacteria, including E. coli, Salmonella, Proteus and Pseudomonas species; as such, there is much interest in the study of Bdellovibrio with a view to its potential use as an alternative antimicrobial therapy. Key to any future research into medical applications of this bacterium is an understanding of its predatory life cycle. Attack phase Bdellovibrio uses rapid flagellar motility through liquid environments to find its prey; upon collision with a suitable prey cell, it forms a strong attachment to the cell and generates a pore in the prey outer membrane. The predator then squeezes through this pore, loses its flagellum and establishes itself in the prey periplasm; the entry pore is then resealed, the prey rapidly killed with its peptidoglycan being modified to produce a rounded, osmotically stable structure termed the bdelloplast. Once safely within the bdelloplast, the Bdellovibrio secretes an arsenal of lytic enzymes into the prey cytoplasm which allow transport of the constituent monomers of DNA, RNA, protein et cetera from the prey back into the growing Bdellovibrio. The predator elongates as a spiral shaped filament within the periplasm of the prey and once the contents of the cytoplasm are exhausted, the filament septates into progeny Bdellovibrio which then re-synthesise flagella and lyse the remains of the prey cell to become free swimming attack phase predators. In this thesis, the role of flagellar motility in the predatory lifestyle of Bdellovibrio was studied through individual insertional inactivation of each of the 6 genes found in the B. bacteriovorus HDIOO genome that encode the flagellar filament protein, FliC. Only one was found to be essential for filament formation yet, contrary to previous hypotheses, this non-motile mutant was still found to be predatory when applied to immobilised prey in a novel fluorescence assay. Thus, the role of Type IV pili in predation was investigated through both microscopic, mutational and transcriptional means. Abolition of the Type IV pilus fibre forming protein, PilA, resulted in a mutant that had no pilus fibres and was incapable of predation using the same assay as in the non-motile FliC mutant strain. Throughout this work, experimental challenges have necessitated the continuing development of electron microscopic techniques for whole cell, flagella, pili and bdelloplast imaging and development of fluorescence methods for use in Bdellovibrio, such as novel use of GFP-encoding transposon mutagenesis and chemical dye-based assays. Work presented here has demonstrated that pili, not flagella, are the bacterial nanomachines required for prey entry and that GFP is a viable tool for future study of this fascinating and potentially therapeutic predator.

Vaccines against infectious diseases are less effective in older adults than in younger adults. This is of significant clinical importance as older adults are also the most vulnerable to contracting, and suffering the most severe consequences of, infectious diseases. Prior research demonstrates that behavioural and psychological factors can modulate the immune system and, in turn, influence how well vaccines work. However, there is a relative paucity of work focusing on older adults. Further, studies have tended to consider only one behavioural or psychological factor at a time - meaning it is unclear which factor, or combination of factors, should be the target of interventions to enhance vaccine effectiveness. This thesis presents three distinct, yet inter-related, pieces of original research which sought to further our understanding of the behavioural and psychological influences on vaccinations and inform the future development of interventions to enhance vaccine effectiveness. First, current approaches to measuring the immune response to vaccination (e.g., ELISAs) are limited in that they require large volumes of sera, antigen, and other consumables. This makes them expensive and time consuming, which limits their utility for larger studies. Thus, the first phase of the research involved establishing a novel high-throughput multiplex antigen microarray assay for quantifying influenza-specific IgG levels in sera. This involved adapting an existing microarray assay and validating it in a series of laboratory experiments. The microarray assay demonstrated acceptable intra- and inter-assay reliability and correlated well with ELISA (H1N1: rho=.534, p < .01; H3N2: rho=.802, p < .00001; B: rho=.454, p < .01). Crucially, the protocols developed could be adapted to suit a variety of research purposes in the future. Second, a prospective longitudinal observational cohort study was conducted to investigate the influence of modifiable psychological and behavioural factors on short and long-term antibody responses to influenza vaccination in older adults (n=138). Diary methods, pedometers, and anthropometric measurements were used to assess nutrition, sleep, physical activity, affect and perceived stress repeatedly over the 2 weeks prior to, and 4 weeks following influenza vaccination. Greater positive affect across the measurement period was found in multivariate regression models to be a significant independent predictor of both short- (β=.189, p=.036) and long-term H1N1 antibody responses following vaccination (β=.296, p=.003): with greater positive affect predicting a more robust antibody response. However, positive affect on the day of vaccination was found be even more salient, explaining greater variance than positive affect over the longer period (H1N1 short term model: R2=.077 vs .058; long term model: R2=.136 vs .125). Greater variability in perceived stress over the measurement period predicted poorer long-term antibody responses to both H1N1 (β=-.268, p < .05) and B strains (β=-.255, p < .05), even after taking overall stress exposure into account. Third, in view of the observed relationship between positive affect and immune responses to vaccination, and in particular the role of positive affect on the day of vaccination, a systematic scoping review was conducted to examine the effects of brief positive mood interventions on immunity. A moderate-sized (31 studies, 38 interventions), but relatively low quality, literature was identified. Few studies included older adults and none examined the effects on in-vivo immune challenges such as vaccination. While there was considerable heterogeneity in the form of interventions that elicited mood improvements, the clear majority of studies reported enhanced immunological outcomes (87.1%). However, many interventions were unsuitable in their current form for implementation within the current already resource-stretched UK health system. Together, the work presented in this thesis points to the potential utility of brief, positive affect enhancing interventions as a way of enhancing immune responses to vaccination in older adults. Further methodologically rigorous research is needed to systematically develop and evaluate positive affect interventions that are both acceptable for older adults and feasible for implementation in the NHS.

The basic model of antigen-presenting cells (APC) activation describes a change of state from resting (immature) to activated (mature) upon encounter with a first signal. However, it is apparent that APC maturation is not binary but instead has a spectrum of outcomes. This likely reflects the complex and dynamic crosstalk resulting from their encounter with a wide variety of stromal cells. A deeper knowledge of these interactions, and how they are processed by APC to instruct immunity is required. This study explores the plasticity of the major APC, namely macrophage (Mφ) and dendritic cells (DC). We test the hypothesis that plasticity in APC function depends on the sequence of signals encountered during generation and activation. Human laboratory models of Mφ and DC were generated from monocytes as follows; monocyte-derived Mφ were cultured with M-CSF (M-Mφ) or GM-CSF (GM-Mφ) and DC with IL-4 and GM-CSF. These were characterised and the impact of cellular (e.g. fibroblasts (FB)) or external factors (e.g. ionising radiation (IR)) studied. Changes in microenvironment considerably alter the functional-phenotype of Mφ however it remains to be shown if their initial polarisation restricts responses. Activation of GM-Mφ elicited IL-12 and -23 whilst M-Mφ expressed IL-10. However, the classic pro-inflammatory signal interferon γ (IFNγ) did not elicit M-Mφ that secreted cytokines to the level expected of GM- Mφ. The M-Mφ displayed mixed phenotype with both pro-inflammatory cytokines (albeit at lower levels, 75% less IL-23, 21% less IL-12) on a background of high IL-10. The IL-23 response of Mφ to re-stimulation was reduced or abrogated upon pre-treatment with LPS but enhanced in Mφ conditioned by IFNγ. These findings demonstrate limitations to Mφ plasticity reflecting initial conditioning and suggest the development of innate memory with a degree of specificity in Mφ responses to external factors. We applied ionising radiation (IR) to our models at doses commonly used for radiotherapy (1-20Gy). Irradiation markedly suppressed IL-23 secretion by both Mφ and DC (p < 0.01) in a dose-dependent manner without affecting viability. This was selective for IL-23 as there was no change in IL-1β, -6, -10, -27 or TNFα, and only a modest decrease in IL-12 secretion. To investigate the mechanism for IL-23 regulation we showed that IR activated Ataxia telangiectasia mutated kinase (ATM) which in turn inhibited the function of cAMP response element-binding protein (CREB) through phosphorylation of an inhibitory residue (S121). ATM activation suppressed IL-23 as inhibition of ATM with KU55933 increased its expression. Furthermore, the small-molecule CREB inhibitor C-Ci reduced IL-23 levels. As a consequence of modulation of IL-23 in APC, IR indirectly affected the generation of T-cell responses, namely Th-17. Cross-talk between APC and stromal cells was modelled by co-culture of DC with FB. The presence of FB prevented IL-23 downregulation by IR. This was dependent on TNFα and IL-1β secretion by DC which were unaffected by IR. In response, FB secreted PGE2 which resulted in a feed-back loop to enforce IL-23 secretion via activation of cAMP. Therefore, whilst IR could be seen to regulate IL-23 in a single cell culture system, this did not occur in co-cultures that simulated the interaction APC have with stroma. In conclusion, we shown that the order in which APC receive conditioning signals ultimately affects how they respond to re-challenge. This is further influenced by a mixture of micro-environmental and external factors. Understanding the molecular pathways through which these cues manifest will result in in vitro models that better reflect the biology of APC and create opportunity for intervention.

Staphylococcus aureus is a highly successful human pathogen capable of colonising and spreading easily between humans while causing a wide range of infections, including life threatening bacteraemias, endocarditis and pneumonia. Virulence gene expression is regulated primarily by the staphylococcal accessory gene regulator (agr) in a cell density-dependent manner termed quorum sensing. Strains of S. aureus, particularly community-associated methicillin resistant S. aureus (CA-MRSA), carrying the genes for a pore forming exotoxin, the Panton Valentine Leukocidin (PVL), have been spreading worldwide over the last 10 to 15 years. As high level production of exotoxins has been implicated in the success of the CA-MRSA clone USA300, a collection of PVL producing clinical isolates from Nottingham were studied with the aim of understanding why PVL production might vary between strains. Sequence typing of the Nottingham strain collection revealed that high PVL producing ST22 agr group 1 and low PVL producing ST30 agr group 3 strains were the most common types. PVL production was stimulated by addition of the type specific exogenous autoinducing peptide (AIP), which activates the agr sensor AgrC, and inhibited by the AgrC antagonist (ala5)AIP-1 if added before a critical cell population density was reached. Analysis of the pvl promoter identified predicted binding sites for RNA polymerase and the SarA protein family of regulators. Promoter pull down experiments confirmed the binding of several staphylococcal regulators to both the agr and pvl promoters including SarA, SarS, Rot and MgrA indicating that these are likely to play a role in the regulation of PVL production. Analysis of the agr locus of high and low PVL producing strains found that a low PVL producing ST22 agr group 1 strain TS13 had a single nucleotide polymorphism (SNP) conferring an N267I substitution in the cytoplasmic domain region of agrC not present in a high PVL producing ST22 strain TS14. Whole genome sequencing of these strains revealed them to be closely related to each other, differing by less than 200 SNPs across their core genomes. While not possessing an SCCmec element, they carried phages encoding pvl and the immune evasion complex (IEC) comprising staphylokinase, the chemotaxis inhibitory protein (CHIPS) and the staphylococcal complement inhibitor (SCIN). AgrC N267I reduced sensitivity to AIP in a bioluminescent reporter for agr which responds to but does not produce AIP. Introducing AgrC N267I into an agr reporter which both produces and senses AIP resulted in delayed autoinduction. This appeared to explain the delay in AIP autoinduction observed in TS13 as compared with TS14 with resulting low PVL production. Other naturally occurring agrC cytoplasmic mutations including T247I and I311T/A343T reduced AIP sensitivity and were associated with delayed autoinduction and reduced exotoxin production. PVL production is closely regulated by agr with expression mediated by several staphylococcal regulators in a cell population density dependent manner. Mutations in agrC occur naturally, delay autoinduction, can have a marked impact on exotoxin production and may be a form of adaptation to niches where high level agr expression is not required.

Aspergillus niger is a black-spored filamentous fungus that forms asexual spores called conidospores (‘conidia’). Germination of conidia, leading to the formation of hyphae, is initiated by conidial swelling and mobilisation of endogenous carbon and energy stores, followed by polarisation and emergence of a hyphal germ tube. These morphological and biochemical changes which define the model of germination have been studied with the aim of understanding how conidia sense and utilise different soluble carbon sources for germination. Microscopy and flow cytometry were used to track the morphological changes and results showed that the germination of A. niger conidia was quicker and more homogenous in rich media than in minimal media. The germination of conidia was also shown to be quicker in the presence of D-glucose than D-xylose. In the absence of a carbohydrate, no visual indicators of germination were evident. Added to this, the metabolism of internal storage compounds was shown to only occur in the presence of a suitable carbon source. Specific environmental carbon sources may therefore serve as triggers of germination, i.e. to initiate the catabolism of stores such as D-trehalose and the swelling of conidia. Studies carried out using D-glucose analogues identified the structural features of sugars that trigger or support conidial germination. These studies showed that the arrangement of atoms on carbons 3 and 4, on the pyranose ring structure of D-glucose, are essential to serve as a trigger of germination. The trigger step preceeds, and is separate from, the energy generation step that supports the continued outgrowth. Transcriptomic studies found that the most significant changes were associated with the breaking of dormancy. The data also revealed that fermentative metabolism present at the early stages of spore germination is rapidly replaced by respiratory metabolism.

Nucleotide excision repair (NER) is a versatile pathway of DNA repair that deals with a variety of DNA lesions, such as UV-induced DNA damage and interstrand crosslinks. In bacteria, the UvrABC system carries out NER. In human cells, XPF and XPG are two structure-specific endonucleases that act in NER. XPF is responsible for a 5' incision at the DNA lesion and XPG carries out the 3' incision. In Archaea, the third domain of life, most species have homologues of some eukaryal NER proteins. Interestingly, Haloferax volcanii encodes homologues of both the eukaryotic NER genes (XPF, XPG, XPB and XPD) and bacterial NER genes (uvrA, uvrB, uvrC and uvrD). In this study, the function of XPG, XPF and UvrA in H. volcanii is investigated. XPG is related to FEN1, a structure-specific 5' flap endonuclease that acts in Okazaki fragment maturation. H. volcanii has a single gene homologous to both XPG and FEN1. The helicase/nuclease hef gene in H. volcanii is the archaeal homologue of human XPF, but also shows homology to Mus81 and FANCM. Mus81 has been found to resolve joint molecules in yeast, while FANCM is required for the repair of interstrand crosslinks in vertebrates. The uvrA gene in H. volcanii is the archaeal homologue of bacterial uvrA, which encodes a protein that plays a vital role in NER at the DNA damage recognition step. This study demonstrates that in H. volcanii, UvrA is involved in the major pathway for repair of UV induced DNA damage. By contrast, Hef and UvrA are involved in two different pathways for the repair of mitomycin C induced DNA crosslinks. Fen1 and Hef have overlapping functions for the repair of DNA cross-links, but not oxidative damage. We also obtain a spontaneous suppressor sfnA, which can suppress the slow growth and MMC sensitivity, but not the UV sensitivity of fen1 deletion mutants. Using plasmid assays, it has been shown that the hef deletion mutant is deficient in accurate end-joining and homologous recombination, including both crossover and non-crossover recombination. In contrast, Fen1 has no significant role in accurate end-joining, but Fen1 may regulate the ratio of non-crossover recombination to crossover recombination.

Colonic inflammation in C. difficile infection is mediated by released toxins A and B. I have investigated responses to C. difficile toxin A and B by primary human colonic myofibroblasts, which represent a distinct subpopulation of mucosal cells that are normally located below the intestinal epithelium and epithelial cell lines, Caco-2 and HT29. Myofibroblasts, isolated from normal human colonic mucosal specimens, Caco-2 and HT29 cells incubated with purified toxin A or B displayed a dose dependent response. Myofibroblast morphology changed to a stellate shaped cell, with processes that were immunoreactive for alpha smooth muscle actin. Most of the myofibroblasts remained viable, with persistent stellate morphology, despite exposure to high concentrations (up to 10 μ g/ml) of toxin A for 72 h. In contrast, a majority of the toxin B exposed myofibroblasts lost their processes prior to cell death over 24-72 h. Investigating toxin A+B on myofibroblasts, at low concentrations, toxin A provided protection against toxin B-induced cell death. Most of the intestinal epithelial, HT29 cells remained viable despite exposure to high concentrations of either toxin (up to 10 μ g/mi). By contrast, a significant loss in cell viability was observed in Caco-2 cells exposed to either toxin. Within 4 h, myofibroblast and epithelial cell types exposed to either toxin A or B lost expression of the non-glucosylated form of Racl, but total intracellular RhoA remained unchanged in myofibroblasts and Caco-2 cells. A time-dependent reduction in RhoA expression was seen in HT29 in response to toxin A or B. Active RhoA expression was lost within 4h in myofibroblasts exposed to either toxin. Despite pre-exposure to high concentrations of toxin A for 3 h, colonic myofibroblasts were able to recover their morphology and proliferative capacity during prolonged culture in medium. This was also shown when pre-exposure to toxin A was extended to 48 h. However, toxin B-pre-exposed myofibroblasts were not able to recover. In conclusion, primary human colonic mucosal myofibroblasts are resistant to toxin A (but not toxin B)-induced cell death. Responses by colonic myoflbroblasts may play an important role in mucosal protection, repair, and regeneration in colitis due to C. difficile infection. Investigation into the apparent resilience of HT29 cells has highlighted the importance of cell specific substrate specificity by C. difficile toxins.

Allergic reactions are pathological immune reactions against common, harmless environmental proteins, termed “allergens”. Type I hypersensitivity reactions are mediated by IgE antibodies; these reactions are the most prevalent, affecting all ages and both genders to almost the same extent. This form of allergic reaction is distributed widely affecting more than 25% of the population in the industrial world. The increased incidence of IgE-mediated allergy is thought to be a combination of environmental and genetic factors, each account for about 50%. The effects associated with this disease may reduce quality of life, daily activities, and in some cases threaten life, hence affecting work and leisure time, increasing healthcare and medication costs. IgE-mediated allergy is defined a complex innate and adoptive immune response to natural harmless environmental allergens in which allergen- specific IgE antibodies are predominantly produced. In addition, it is believed that other immunoglobulin isotypes might be involved in atopic diseases (mainly IgG4), although the roles of these antibody isotype are still debatable whether they are pathogenic or protective antibodies. Generally, diagnostic procedures for IgE-mediated allergy begin with the patient’s history, clinical symptoms, physical examination, in vivo provocation tests, and are eventually confirmed with laboratory investigations. In this thesis we develop a microarray technology for use in-house for comprehensive investigations of immunoglobulin reactivity profiles against numerous allergens in atopic disease. Several printing and blocking buffers were developed in-house to preserve functionality of the arrayed proteins, minimise background noise, and increase signal intensity. The new diagnostic platform was validated by set of quality controls, then applied on large-scale applications (allergic patients, general populations, suspected asthmatics and healthy control individuals) to measure total and specific IgE antibodies. Application of the developed antibody microarray platform identified the difficulty of setting cutoffs using a “healthy control group”. We tested both the 75th and the 95th percentiles of the healthy control group as a cut-off point for total IgE levels that resulted in 39.3ng/ml and 70.68ng/ml respectively. When the 75th percentile was selected, the number of participants in the allergy patients classed as positive above the cutoff points was 92 out of 105 (87.6%). However, if the 95th percentile was selected, the number of positive patients was 63 out of 105 (60%) in the same group. On the other hand, the developed allergen microarray platform was used to immunoprofile allergen specific immunoglobulin subtypes in the four different cohorts against 70 SPT allergens. MeV software was used to present the reactivity profiles of the immunoglobulin subtypes semi-quantitatively that resulted in the hierarchical clustering according to allergen species. The reactivity profiles of allergen specific IgE in the allergic group recognized multiple allergens while specific IgE in the healthy control group was negative except for one serum sample. Specific IgE in the random population was lower than in the suspected asthmatics although it recognized allergens from different species (ingested, inhaled, venom). Similarly, this multi-allergen microarray platform was adapted to investigate the roles of other antibody isotypes (IgG1, IgG2, IgG3, IgG4, IgA1, IgA2) in IgE- mediated allergy. The study revealed no roles for these isotype except IgG4 that showed multiple recognition patterns in serum samples of healthy individuals and allergic patients. A comparison between microarray results and provocation testing (SPT) was performed for the samples from suspected asthmatics. To enable direct comparison, IgE:IgG4 ratio was utilized as IgE levels alone would not be expected to correlate with wheal and flare reaction. The IgE:IgG4 ratio showed good correlation with the six allergens tested by SPT (r=0.9299, p<0.0001). In addition, the multi-allergen chip can detect additional responses outside the standard six SPT- allergens performed routinely, and moreover reports on IgG4 responses that are never assessed by SPT, but are potentially important to the likelihood of asthma and allergy. Overall, the results indicated that the new developed microarray platform is highly specific, sensitive and reproducible. Subsequently, this high-throughput microarray platform can work as a diagnostic tool for monitoring the development of allergic diseases, predicting allergic reactions particularly for polysensitised patients, evaluating patient response to treatments, and eventually allowing the classification of allergic individuals. In conclusion, a new high-throughput multi-allergen platform has been developed that will certainly improve the future of allergy diagnosis.

Neisseria meningitidis (Nm) can cause life-threatening bacterial meningitis and septicaemia. The high mortality of meningitis is associated with meningococcal invasion of the central nervous system (CNS) by breaching of the blood-brain barrier (BBB). Nm elaborates several cell-surface and secreted virulence factors that promote host invasion and colonisation. One important class are the autotransporter proteins. Adhesion and penetration protein (App), meningococcal serine protease A (MspA), Immunoglobulin A1 protease (IgA1p) and Neisserial autotransporter lipoprotein (NalP) are serine protease autotransporters of Nm; all have previously been demonstrated to have roles in meningococcal virulence. For App and MspA, uptake by host cells, nuclear localisation, histone clipping and induction of apoptosis have been described in our laboratory. However, a time course for the efficient uptake and nuclear localisation of these proteins into human brain microvascular endothelial cells (HBMECs) and details regarding the mechanism for histone clipping remained unknown. To address these points, both App and MspA and proteolytically-inactive mutant derivatives were expressed using the pColdTF vector system and purified under native conditions. Using confocal scanning microscopy optimal uptake and nuclear localisation of recombinant fusion proteins containing the functional passenger domains of both App and MspA in HBMECs was shown to occur 8 h-post exposure. Furthermore, the requirement for the active site serine residue in both autotransporters for H3.1 clipping was demonstrated, and human coagulation factor V (FV) was shown to be an additional substrate for both proteins. NalP is a cell-surface maturation protease which processes App, MspA and other meningococcal surface proteins such as IgA1P, LbpB and NhbA, and thus modulates the cell surface and secretome of the organism. Previous studies aimed at functional characterisation of NalP have typically relied on the phenotypic comparison of wild-type and nalP-mutant derivatives. Here an active recombinant NalP was expressed and purified, and used to investigate the interaction of NalP with host cells in order to more comprehensively elucidate the role of NalP during meningococcal interaction with the host. A recombinant NalP (rNalP) passenger domain was purified under non-denaturing conditions using immobilized nickel chromatography. Although rNalP had apparent molecular weight 8-10 kDa less than that of NalP secreted by wild-type meningococci, it was functional as determined by its ability to process human complement 3 (C3). rNalP was shown to cleave human coagulation factor V (FV), a proteolytic event which is likely to contribute to bacterial pathogenesis. Binding and uptake of rNalP into human cells was demonstrated by flow cytometry and confocal microscopy. Interestingly, rNalP was differentially localised to different cellular compartments in different cell types. Treatment of HBMECs with rNalP resulted in increased levels of IL-6 and IL-8, and decreased levels of TNF-α in culture supernatants. rNalP was shown to clip histone 2B but not other histone proteins. Using a re-tagging approach a number of rNalP-interacting host proteins were identified. These included proteins of the candidate membrane, cytoplasm, cytoskeleton endoplasmic reticulum, mitochondrion, and nucleus. Some of these proteins are likely to be involved in trafficking of NalP and the cellular response to this protein. Overall, the findings of this study expand our knowledge on the contribution of three autotransporters to meningococci pathogenesis and provide a platform to further explore the host response to NalP uptake and the epigenetic changes associated with autotransporter host interactions, which may guide the development of future therapeutic interventions.