Magister Public Health - MPH / Immunisation is considered to be amongst the most successful and cost-effective disease prevention interventions available. The Expanded Programme on Immunisation (EPI) in Namibia was established in 1990 to ensure that the immunisation of children takes place within the prescribed age frame. However, continued measles outbreaks, particularly in the Kavango region, are evidence of poor EPI progress, with vaccination coverage being below80% per district. The reasons for the low uptake of measles immunisation in the Nyangana district in the Kavango region are not clearly understood. The aim of this study was, therefore, to investigate the perception of mothers/caregivers of factors that impact on the uptake of measles immunisation in the Nyangana Health District, with a view to improving measles immunisation coverage. Methodology: A qualitative exploratory study design was used to collect data from the study participants. In-depth interviews were conducted with 10 mothers of children under 5 years of age, for both children who received, and those who did not receive measles vaccination. Data was audio-taped and transcribed verbatim. The recorded interviews were translated from the Gciriku language to English. Data was analysed through the use of the Thematic Content Analysis approach. The transcribed interviews and narratives from the research assistant’s notes were organised into codes, sub-themes and main themes. In the final phase, themes were integrated and interpreted, by identifying facilitating factors for those who took their children for immunisation, and barriers for those who did not take theirs. The researcher facilitated assistance to children who did not receive their measles dose, to receive it. Ethical requirements were adhered to throughout the research study process. Results: The study showed that mothers had both positive and negative perceptions about immunisation. The findings revealed that information, and past experience of measles ,irrespective of the level of education, support from a spouse or family members, availabilityof services and convenience of time schedules, increased the uptake of immunisation on thepart of mothers/caregivers. However, it also emerged that supply-side factors, such as lack of information sharing between health care providers and mothers, hindered effective communication. Additionally, inconvenient time schedules and time constraints, staffshortages, health care providers’ attitudes, inaccurate data being kept of children immunisedat other health facilities, inadequate outreach services and perceived lack of supervision in the health facilties all contributed to the low uptake of immunisation. Demand-side factors that affected the uptake of immunisation included: socio-economic constraints that led to an inability to pay transport costs to access immunisation services; lack of support from a spouse; other family members and other support structures in the community also impacted on immunisation uptake, despite the reported awareness and willingness to use immunisation services. Conclusions and recommendations: The study concludes that the relationship between health care providers and mothers/caregivers, and support from other social structures, should be good, in order to motivate mothers to use immunisation services. The study recommends that the following aspects be addressed, as they have the potential to improve the low uptake of measles immunisation: patient/provider relationship, information sharing, and supervision in the health facility, access to services, availability of outreach services, improved data tracking and active involvement of all stakeholders. Laziness was overwhelmingly offered as an explanation for missing measles immunisation, although there are suggestions that there might be underlying causes for what is perceived as laziness, which require further exploration, especially in terms of socio-cultural barriers to immunisation. It is recommended that an in-depth look at the perceptions of health care providers and key informants should be conducted to search for further understanding of contributing factors.

Contexte : La vaccination est la mesure la plus efficace pour prévenir l'influenza et les infections à pneumocoque qui sont des sources importantes de morbidité et mortalité. Au Québec, les couvertures vaccinales contre ces deux infections n'atteignent toutefois pas l'objectif souhaité de 80%. Les groupes de médecine de famille (GMF) grâce à certains de leurs attributs (responsabilité envers la clientèle inscrite, présence d'infirmières et appui technique informatique) semblent particulièrement prometteurs pour favoriser la mise en place de services d'immunisation efficaces. Un modèle idéal d'organisation des services d'immunisation a été proposé en tenant compte des stratégies reconnues efficaces pour augmenter les couvertures vaccinales, des pratiques pour assurer la qualité de l'acte vaccinal et de l'organisation des GMF. Objectifs de l'étude. Apprécier, auprès des médecins, la faisabilité d'un modèle d'organisation des services d'immunisation contre l'influenza et le pneumocoque dans les GMF. Méthode. Une étude descriptive à deux volets a été réalisée en 2005-2006 auprès des médecins responsables des 19 GMF de la Montérégie. Le volet 1, soit des entrevues téléphoniques avec les médecins responsables ou autres informateurs clés désignés par eux, a permis de connaître certaines caractéristiques des GMF, leurs pratiques d'immunisation ainsi que les ressources et les normes d'encadrement qui y sont associées. Des analyses descriptives ont été faites pour décrire les pratiques actuelles d'immunisation contre l'influenza et le pneumocoque dans les GMF et apprécier les écarts entre les pratiques présentes et le modèle proposé. Par la suite, les entrevues face à face du volet 2 avec les médecins responsables, nous ont fait connaître l'opinion de ces derniers sur les facteurs favorables et obstacles par rapport au modèle proposé. L'analyse qualitative de ces facteurs a identifié les conditions requises pour implanter ce modèle. Résultats. Volet 1: Les pratiques d'immunisation contre l'influenza et contre le pneumocoque varient beaucoup d'un GMF à l'autre et même entre les sites d'un même GMF. Néanmoins de façon globale, les services de vaccination contre l'influenza correspondent assez bien au modèle proposé. L'immunisation contre le pneumocoque est cependant moins bien organisée. Par ailleurs l'écart le plus remarquable entre les pratiques présentes dans les GMF et le modèle proposé est l'absence de stratégie s'adressant aux vaccinateurs. Volet 2: L'analyse de l'opinion des médecins permet de faire certains constats par rapport aux attributs qui nous semblaient a priori favorables à la mise en place de pratiques vaccinales efficaces. Pour ce qui est de la responsabilité envers leur clientèle inscrite, bien que les médecins des GMF se sentent responsables et accordent une grande importance à l'immunisation, l'inscription de la clientèle n'est pas actuellement utilisée pour la leur offrir. La présence d'infirmières est une condition moins favorable qu'anticipée, car les tâches de ces dernières limitent leur disponibilité pour l'immunisation. Finalement l'appui technique informatique tarde à venir, les médecins responsables croient que des logiciels performants pourraient faciliter la mise en place de plusieurs stratégies et leur permettre d'être plus efficaces dans leurs pratiques vaccinales. Conclusion. Cette étude a permis de tracer le premier portrait détaillé des services d'immunisation contre l'influenza et le pneumocoque dans les GMF. Les médecins sont motivés et aimeraient améliorer leurs pratiques vaccinales. Pour le faire, ils ont besoin de ressources: un système d'information performant, certains outils, une rémunération.

Les infections causées par Staphylococcus aureus sont associées à un fort taux de mortalité au sein de la société. En plus d'être responsable de diverses infections, cette bactérie possède une panoplie de résistances aux antibiotiques ce qui rend son traitement beaucoup plus difficile. La vaccination semble être la voie la plus prometteuse compte tenu de l'inefficacité des antibiotiques. Bien que cette bactérie cause beaucoup de dommages, il est difficile d'enrayer toute source de S. aureus puisque celle-ci est un commensal de l'humain et se retrouve au niveau du nez et sur la peau de porteurs sains. L'étude de l'efficacité d'un vaccin suite à une infection septicémique avec différentes souches de S. aureus chez la souris était le principal objectif du présent projet. Il s'agissait donc de comparer la protection immunitaire engendrée par le vaccin face à ces souches. Les souches SHY97-3906 (isolat de mammite), MRSA256c (isolat nasal chez l'humain) et Newman (souche contrôle séquencée, isolée au départ d'une ostéomyélite) ont été utilisées. Le vaccin que nous avons employé est composé de quatre protéines différentes hautement conservées chez les différentes souches de S. aureus soit; IsdB (iron-regulated surface determinant ), HarA/IsdH (haptoglobin receptor A ), CIfA (clumping factor A ) et GapC/B (glycéraldéhyde-3-phosphate déshydrogénase). De plus, deux adjuvants ont été utilisés, le PCEP (polyphosphazene ) et le pGM-CSF (granulocyte-macrophage colony-stimulating factor plasmidique). Ces différents éléments de notre vaccin ont été choisis suite à des résultats préliminaires obtenus au laboratoire par Marie Rivest. Ce vaccin a donc été testé chez la souris (CD-1) face à une infection septicémique. Suite à des résultats non concluants, l'objectif de l'étude s'est plutôt redirigé vers l'investigation de l'échec du vaccin face à la septicémie. Puisque le vaccin ne semblait pas pouvoir contenir l'infection, il était important de vérifier l'efficacité des anticorps à lier la bactérie. Une grande variabilité a été obtenue ce qui a poussé l'analyse plus loin, au niveau de l'expression des gènes cibles. Des PCR réguliers et quantitatifs ont été réalisés sur les différentes souches afin de déterminer les différences d'expression. De plus, lors de mise au point du modèle de septicémie des différences notables de virulence in vivo avaient été observées entre les souches de S. aureus . Un lien a donc pu être fait entre l'habileté des anticorps à lier les souches, l'expression des protéines cibles et la virulence des souches afin d'expliquer en partie, l'échec de notre vaccin face à S. aureus . Comme il a été démontré dans la littérature que les individus portant S. aureus au niveau du nez étaient plus enclins à développer des infections, une mise au point du modèle nasal a été effectuée. Cependant, aucun résultat significatif au niveau de la protection n'a été obtenu.

Titre. Évaluation de la couverture vaccinale des jeunes enfants de la Montérégie au regard des facteurs sociodémographiques et impact de l'ajout des nouveaux vaccins. Contexte. En Montérégie comme ailleurs au Québec, les couvertures vaccinales des enfants de 2 ans selon les facteurs sociodémographiques sont peu connues. Avec l'ajout des nouveaux vaccins, le calendrier vaccinal des jeunes enfants se complexifie davantage ; récemment, le vaccin contre le pneumocoque, l'influenza et la varicelle ont été introduits pour la vaccination des nourrissons. On anticipait que cet ajout ait un impact sur le respect du calendrier à cause entre autres de l'administration de multiples injections lors d'une même visite. La présente étude vise à évaluer la couverture vaccinale des enfants de la Montérégie au regard des facteurs sociodémographiques et à explorer l'impact de l'ajout des nouveaux vaccins sur l'observance du calendrier de vaccination et l'opinion des parents. Méthodologie. Étude descriptive transversale par questionnaire postal auto-administré auprès d'un échantillon aléatoire stratifié de deux cohortes d'enfants tirés du Fichier des naissances de la Montérégie : 1) cohorte 2002-03 : enfants nés entre le 1er mai 2002 et le 30 avril 2003 (ancien calendrier) et 2) cohorte 2004-05 : enfants nés entre le 1er mai 2004 et le 30 avril 2005 (nouveau calendrier). Deux enquêtes ont été réalisées : Enquête 1 (automne 2005-hiver 2006) effectuée auprès des parents pour lesquels l'adresse a pu être validée par Canada 411 (taux de reponse=56 %) ; enquête 2 (automne 2006) effectuée auprès des parents dont l'adresse a été validée par la RAMQ (taux de réponse=57 %). Les données vaccinales incomplètes ou manquantes ont été complétées par vérification des dossiers vaccinaux chez le vaccinateur. Les variables dépendantes à l'étude sont le statut vaccinal, l'observance au calendrier et l'opinion des parents sur les injections multiples. Les variables indépendantes retenues sont les caractéristiques sociodémographiques. Résultats. Les résultats montrent que 77 % (IC 95 % : 74,8 % - 80,0 %) des enfants de la cohorte 2002-03 ont reçu tous les vaccins requis et 2 % n'ont jamais été vaccinés. La couverture vaccinale varie au palier local entre 73 % et 95 %. Parmi les facteurs sociodémographiques étudiés (âge du père et de la mère, scolarité du père et de la mère, statut marital, rang de l'enfant dans la fratrie, taille de la famille, poids à la naissance et statut socioéconomique), seul le rang de l'enfant dans la fratrie est associé significativement au statut vaccinal de l'enfant. Une plus forte proportion d'enfants à un statut vaccinal complet parmi les premiers nés de la famille (p=0,037). L'analyse de l'observance du calendrier par l'approche de survie de Kaplan-Meier montre un retard dans la réception de la deuxième et la troisième dose du vaccin pentavalent DCaT-P-Hib des enfants de la cohorte 2004-05 comparés à ceux de la cohorte 2002-03 (Nombre de jours médians de retard : 2 jours sur le 2e vaccin (p=0,013) et 4 jours sur le 3e vaccin (p<0,001)). La majorité des parents (76 %) préfèrent 2 à 3 injections lors d'une même visite pour leur enfant et 61 % craignent une augmentation du risque d'effets secondaires lorsque plusieurs vaccins sont administrés à la même visite. Conclusion. Bien que les retards de vaccination constatés semblent moins importants qu'anticipés, la couverture vaccinale observée en Montérégie s'écarte des objectifs provinciaux. Des efforts importants devront continuer à être effectués pour obtenir une meilleure couverture vaccinale. Une attention particulière devra être accordée aux enfants derniers nés de familles nombreuses et à la sensibilisation des parents pour réduire les fausses croyances notamment sur la crainte d'effets secondaires accrus.

La transfusion sanguine est un traitement essentiel à la survie de millions de patients. Son principal risque immunologique est l’allo-immunisation post-transfusionnelle. Elle se traduit par la production d’allo-anticorps contre des antigènes de globules rouges (GR) conduisant à des hémolyses post-transfusionnelles. Les mécanismes à l’origine de la tolérance des GR ou de son inhibition lors de l’allo-immunisation sont mal connus. Ainsi, mes travaux de thèse, portant sur la compréhension de ces effets, se sont articulés en trois parties avec 1/ l’étude des conditions optimales aux réponses allo-immunes, 2/ l’étude des effets d’une stratégie thérapeutique utilisant un anticorps monoclonal et 3/ l’étude des effets immunomodulateurs, incluant la tolérance, médiée par des composants présents dans les concentrés de globules rouges (CGR).Afin d’étudier l’allo-immunisation, nous avons utilisé le modèle murin. Nous montrons qu’une variation du délai entre la transfusion et la stimulation du TLR3 impacte la réponse immune dans la rate. Une activation importante des lymphocytes T CD4+ (LT CD4+) allo-réactifs accompagnée d’une production accrue d’allo-anticorps ont été montrées à 7 jours de délai. Afin de limiter l’allo-immunisation, l’utilisation d’un anticorps anti-CD20 déplétant les lymphocytes B (LB) montre une altération des LB mais surtout des LT CD4+ impliqués dans le processus d’induction de l’allo-immunisation. Enfin, la modification du phénotype des cellules dendritiques CD11c+ de la rate des souris transfusées, observée hors contexte inflammatoire, suggère une maturation incomplète à l’origine d’une tolérance antigénique. Pour finir, l’analyse de différents composants présents dans les CGR confirme l’existence de microparticules (MPs) lymphocytaires. Ces MPs présentent des molécules inhibitrices et pourraient donc être impliquées dans la tolérance des antigènes transfusés.En conclusion, mes travaux montrent la coopération des DCs avec les LT CD4+ permettant celle des LT CD4+ avec les LB pour induire une réponse immune. Comme toute réponse humorale, nous confirmons que l’allo-immunisation fait intervenir des DCs, des LT CD4+ et des LB. Ces résultats ouvrent de nouvelles voies de recherche pour mieux caractériser l’allo-immunisation en particulier chez les patients drépanocytaires qui sont les plus touchés. / Red blood cell (RBC) transfusion is a life-saving treatment for millions of patients. However, its main immunological risk is RBC alloimmunization resulting in antibody production against RBC antigen. Alloimmunization can lead to severe complications threatening the life of the patient. The mechanisms explaining RBC alloimmunization are poorly understood. Therefore, my doctoral work aiming at understanding transfusion effects, was subdivided into three parts with 1/ the study of optimal conditions for alloimmune responses, 2/ the impact of a therapeutic strategy using a monoclonal antibody to inhibit alloimmunization and 3/ the study of immunomodulatory effects of transfusion, including tolerance, through components present in the RBC concentrates.We also used HOD murine model for the study of alloimmunization to show an impact of the delay between the TLR3 agonist injection and the transfusion on immune responses against RBCs. At 7 days of delay, we have demonstrated an important alloreactive CD4+ T-cell activation and a wider alloantibody production. Furthermore, B-cell depletion, using a monoclonal anti-CD20 antibody, revealed potential changes on LB implicated in alloimmunization induction and mostly on alloreactive CD4+ T cells. We finally observed a modification of splenic CD11c+ DC phenotype from transfused mice out of a TLR context. This suggest an incomplete maturation that could explain antigen-specific tolerance. The investigation of several components in RBC concentrates confirmed the presence of microparticules (MPs) issued from T lymphocytes. These MPs carry inhibitory markers and could thus inhibit DC activation to induce antigen-specific tolerance.Therefore, my doctoral work highlights the implication and the cooperation of DCs with CD4+ T cells to allow cellular cooperation between CD4+ T cells and B cells for immune response induction. As in any humoral response, we confirmed that RBC alloimmunization involves DCs, CD4+ T cells and B cells. In addition, these results are opening up new areas of investigation for a better characterization of alloimmunization in particular in the most affected patients, the SCD patients.

Throughout history a diversity of animal species have been used and studied extensively in the development of vaccines for the benefit of humans and animals alike. As mice are a relatively easy species to maintain, handle and manipulate, and have the advantage of being cost effective, they are commonly employed as animal models in the investigation of immunisation strategies against mucosal associated pathogens. Vaccine research against the human gastric pathogen Helicobacter pylori is extensively conducted in a mouse model and typically uses intra-gastric administration for the testing of potential vaccine candidates. An inherent complication with this route, however, is that the vaccine constituents may be inadequately delivered to sites of specific immunity and consequently may not be the optimal method for vaccine delivery. In the present study a mouse model of H. pylori infection was used to determine the efficacy of alternate mucosal routes of immunisation from examination of protective immunity, immune responses and the practical aspects of vaccine administration. Commencing with the optimisation of intra-intestinal immunisation, the direct injection of a H. pylori vaccine to initiator sites of the mucosal immune system established baseline data of dose rates for the comparative analysis of intra-gastric, intra-nasal and intra-rectal immunisation. Following the development of simple administration techniques whilst maintaining the welfare of the animals, intra-nasal immunisation was shown to elicit the highest level of prophylaxis against H. pylori challenge. Effective prophylaxis was also shown to be dependent upon a specific ratio of the vaccine constituents. When using whole cell lysate of H. pylori and the mucosal adjuvant cholera toxin, the ratio of antigen:adjuvant for optimal protective immunity was 10:1. The outcomes of this study have proved conclusively the necessity for optimisation of all aspects of immunisation in an animal model of infection. / Master of Science (Hons)

Acute otitis media is the most common illness diagnosed during early childhood that can cause significant morbidity (Brook, 1994) and sometimes can cause irreversible sequelae such as a hearing defect and subsequent learning difficulties (Klein, 1994). The aims of the research presented here were to study some aspects of the middle ear defence mechanisms in both immune and non-immune rats following experimental otitis media (OM) with two pathogens nontypeable Haemophilus influenzae (NTHi) and Moraxella catarrhalis (M. catarrhalis). This study also aimed at developing a suitable technique for preparing immunohistochemical staining of middle ear sections (chapter 2). A previous study has shown that a regime where rats received an IPP immunisation combined with an IT boost was effective in enhancing clearance of a middle ear infection with the same strain of NTHi and also in the presence of a concomitant viral infection (Moore et al, 2001). Results of this study have shown that for NTHi infection a distinct cellular influx to the middle ear in the immune rats was accompanied by an enhanced bacterial clearance compared to the non-immunised rats (chapter 3). This cellular influx was responsible for the remarkable reduction in the bacterial number. The sharp decline in PMNs numbers in the NTHi immunised rats that followed complete bacterial clearance at 72h post infection (Table 3.1) indicate a more effectively controlled down regulation of this cell infiltrate than the non-immunised rats. For M. catarrhalis infection, there was no difference in cell infiltrate between immune and non-immune rats, but enhanced clearance of the bacteria were observed for the immune animals. The histopathological changes in the middle ear mucosa of rats with experimentally induced infection were studied to provide a better understanding about the distribution of the inflammatory cells and changes in the mucosa during the first 24h post challenge with NTHi and M. catarrhalis (Chapter 4). These changes have not been previously studied for the two pathogens at 24h post challenge in rats. Induced infections with the two pathogens were found to produce similar histopathological changes but more inflammatory infiltration was observed within the infected mucosa with NTHi than that seen with M. catarrhalis. The infections were characterized by increased thickness of the middle ear mucosa, Eustachian tube mucosa, periosteum and tympanic membrane. There was also an increase in the number and size of small blood vessels at all sites, and these small blood vessels seem to be the source of the inflammatory infiltration into the middle ear mucosa and middle ear cavity during the infection. These findings provided an essential background to the immunohistochemical study. The effect of mucosal immunisation on the distribution of CD4+T cells and CD8+T cells has not been investigated previously. Results of the present study (Chapter 5) show the pattern of distribution of these cells during the first 48h post infection with NTHi in the rat. The number of CD4+and CD8+T cells peaked at 24h post infection in the nonimmunised animal and were highest at 48h post-infection in the immunised rats. The difference in response in the immunised rats may represent regulation of the inflammatory response by the immune system. The inflammatory response regulation is indicated by the difference in cellular influx into the immune rats and the response in the immune rats that corresponds to enhanced bacterial clearance prior to a decrease in numbers of inflammatory cells once the bacteria was no longer detected (Chapter 3). This resolution of the inflammatory mass would reduce the opportunity for continued damage to local tissue. These changes are also supported by the reduction in the thickness of the middle ear mucosa of the immunised rats especially at 24h and 48h post-infection (Chapter 5). This study has shown that there are distinct differences in the rate of bacterial clearance and cellular changes in the middle ear mucosa and tympanic bulla in immunised rats during a middle ear infection. Future studies are still required to gain a better understanding of differences in the inflammatory response for both pathogens, NTHi and M. catarrhalis.

Heterogeneity in immunodominant outer membrane proteins has been proposed as a significant factor in the failure of an NTHi infection to induce immune protection against subsequent infections. This study has examined the vaccine potential of three outer membrane proteins in an attempt to identify conserved regions that could be targeted by an immune response after vaccination. The three proteins investigated were: TbpB, P5 and P48 (HI0164). The optimal route of immunisation in clearing a bolus inoculum of NTHi to the lung in the rat has been shown to be a combination of gut sensitisation with a respiratory boost and this regime was used in the present study. A panel of NTHi isolates was assessed to determine the frequency with which strains were able to bind transferrin and thus be targeted by a TbpBspecific immune response. A high proportion of strains was able to bind transferrin with similar frequencies in isolates associated with infection and those from normal throat swabs. A protocol was developed to purify nonlipidated recombinant TbpB from NTHi using a glutathione-Stransferase (GST)-rTbpB fusion protein and Glutathione-Sepharose affinity chromatography. Mucosal-directed immunisation with rTbpB significantly enhanced clearance of an NTHi challenge to the lung, however, whilst rTbpB-specific antibodies were cross-reactive on Western immunoblots, the cross-reactivity was variable in both transferrin binding inhibition assays and bactericidal activity. This suggested that the rTbpB-specific humoral response would be variable in the recognition of heterologous NTHi isolates. The secondary structure of P5 has been controversial with several reports suggesting that P5 was a fimbrin protein composed of coiled coils. In this present study the interstrain variation in P5 amongst isolates from diverse anatomical sites, as well as computer prediction methods and spectrophotometric analysis, generated a model of P5 based on the homologous E. coli protein, OmpA. This model suggested a B-barrel conformation with no evidence of coiled coils. Synthetic peptides corresponding to conserved regions of P5 that were thought to be surface exposed, as well as a region (H3) with some homology to a protective epitope in the P. aeruginosa protein, OprF, were then combined with a "promiscuous" T cell epitope from the measles virus F protein (MVF) and used for immunisation studies. Whilst variable protection was seen with the peptides, the MVF/H3 peptide was the most efficacious of the antigens assessed in this study in enhancing clearance of NTHi. This occurred in the absence of detectable peptide- or PS-specific antibody leading to the suggestion that cell mediated responses may have played an important role in enhancing clearance in this model. The highly conserved nature of the region in P5 represented by the H3 peptide suggests that further study should be focused on this peptide as a potential NTHi vaccine candidate. The last antigen, P48, is homologous to a A. pleuropneumoniae antigen, AopA, which has been proposed to have potential as a vaccine component against pleuropneumonia in pigs. Sequence analysis of the gene encoding P48 from several isolates showed that this protein was well conserved. Recombinant P48 was purified from a GST-rP48 fusion protein and used for immunisation, which also conferred significant protection. However, immunisation with rP48 was not as efficacious as immunisation with the MVF/H3 peptide. Whilst immunisation with rP48 induced high antibody titres, no bactericidal activity could be detected indicating that bactericidal antibody had not contributed to the observed clearance. In addition, the rP48- specific serum IgG was predominantly of the IgG2a isotype suggesting that Thl cell mediated responses had been induced by immunisation with rP48.

One of the major challenges of vaccine development is the conservation of immunogenicity and protective efficacy through the stages of design, production, formulation and delivery. The critical issue is that how and in what form an antigen is taken up by antigen presenting cells for proteolytic processing and presentation to the immune system bound to MHC can have dramatic effects on the activation of Th cells to drive clonal responses and induction of immunological memory. Nontypeable Haemophilus influenzae (NTHi) is a pathogenic commensal of the human respiratory tract that causes diseases with enormous socioeoconomic burdens. There is no licensed vaccine, although the potential for vaccination with outer membrane components to reduce the incidence of disease caused by NTHi has recently been demonstrated in clinical trials. The issue of immunomodulation was explored in this thesis in the context of the further evaluation of a leading NTHi vaccine candidate, the outer membrane protein OMP26. The efficacy of recombinant OMP26 (rOMP26) against NTHi challenge has been previously demonstrated in mice, rats and chinchillas. In rats, efficacy was shown to be restricted to the precursor form (containing the signal peptide) and not the mature form of rOMP26. The immunodulatory effects of changes to the rOMP26 structure were further investigated in this thesis. A range of structural variants of rOMP26 were constructed in view of reducing extraneous plasmid-derived sequence from the antigen and to introduce a unique cysteine residue as a potential conjugate site for multivalent vaccine development (Chapter 2). It was demonstrated that minor structural changes to rOMP26 such as the addition, deletion, modification or relative positioning of a single amino acid or bulky group, designed to increase the efficiency of production or introduce (cysteine) conjugation sites, altered the expression of the protein in E. coli and the immunogenicity in Balb/C mice. Furthermore, in contradiction to the published report (El-Adhami et al. 1999) and a new study in rats (Chapter 3), there was no positive effect of the signal peptide in mice, with precursor and mature forms of rOMP26 equally immunogenic (Chapter 2). Following confirmation of the need to retain the signal peptide for the immunogenicity of rOMP26 in rats, a precursor form (rOMP26VTAL) in which the conserved n-region of the signal peptide was deleted, and shown to reduce the efficiency of the cleavage of the signal peptide by signal peptidase during protein overexpression in E. coli (Chapter 3). Not only did this deletion result in an increase the yield and stability of the purified precursor protein, but rOMP26VTAL was highly immunogenic and enhanced the clearance of NTHi from the lungs of challenged rats. The potential for signal peptides to be exploited as an immune-enhancing moiety in a proteinaceous vaccine is discussed. Following the development of rOMP26VTAL as a production optimised variant of rOMP26, the next step was to test the feasibility of rOMP26VTAL as a component of a multivalent vaccine (Chapter 4). Two chimeras were constructed with LB1(f)2,1,3, a trivalent synthetic B-cell epitope from the extracellular loop 3 region of the P5 fimbrin protein of NTHi, positioned at the N- or C-terminus of rOMP26VTAL. The solubility of rOMP26VTAL was affected by the fusion, with both chimera constructs expressed only in the insoluble fraction, thus requiring a denaturing protocol for purification. Although rLB1(f)2,1,3-OMP26VTAL was expressed and purified as a more stable protein and in greater yield than rOMP26VTAL-LB1(f)2,1,3, the relative positioning of the fusion was important and rOMP26VTAL-LB1(f)2,1,3 was significantly more immunogenic in rats than rLB1(f)2,1,3-OMP26VTAL. In addition, rOMP26VTALLB1( f)2,1,3, but not rLB1(f)2,1,3-OMP26VTAL induced a significant degree of bacterial clearance following pulmonary challenge with NTHi, in levels comparable to the highly efficacious rOMP26VTAL construct. In the third part of the thesis, bacterial ghosts were evaluated as a novel mucosal delivery technology for rOMP26VTAL and rOMP26VTAL-LB1(f)2,1,3, (Chapter 5). To mimic the natural presentation of OMP26 and P5 fimbrin antigens on the cell surface of NTHi, an OmpA� sandwich fusion surface display system was developed for the outer membrane expression of the OMP26 constructs in E. coli ghosts. Following gut immunisation, but not intranasal immunisation even when co-administered with the cholera toxin�derived adjuvant CTA1-DD, bacterial ghosts were successful at presenting OMP26VTAL and rOMP26VTAL-LB1(f)2,1,3 to the immune system for the induction of enhanced clearance of NTHi in the rat pulmonary challenge model. Although this study was the first to demonstrate enhanced bacterial clearance induced by heterologous antigens expressed in the outer membrane of bacterial ghosts, future studies with ghosts will require optimisation of the expression levels of the OmpA� fusion proteins possibly to avoid cross-reactive responses related to high doses of ghosts in the inoculum. This thesis presents data that both supports the further evaluation of rOMP26 constructs for clinical trials, and has demonstrated the significant effects of structural changes, method of production and delivery system can have on the immunogenicity of a candidate vaccine. Such knowledge will contribute to and provide some new approaches for enhancing the efficiency of vaccine development against a range of diseases including those caused by NTHi. Major Outcomes: 1. Demonstration that the immunogenicity of rOMP26 antigen constructs is affected by structural modifications and their positioning within the construct, and by the delivery system. 2. Development of rOMP26VTAL, an rOMP26 construct with the KNIAK sequence deletion of the signal peptide n-region. This protein retains the immunogenicity and protective efficacy of rOMP26, but is produced with reduced cleavage of the signal peptide, resulting in higher yields and a stable protein. Lacks extraneous plasmidderived multiple cloning site sequence, and is produced in high yield as a stable protein. 3. Construction of a NTHi rOMP26VTAL-LB1(f)2,1,3 chimera antigen that induced enhanced clearance of NTHi in an acute pulmonary challenge model in rats. 4. Development of an OmpA� surface display system for the expression of rOMP26 antigen constructs in the outer membrane of E. coli/bacterial ghosts 5. Bacterial ghosts were successful as delivery vehicles for rOMP26 candidate vaccine constructs when delivered in the gut.

Since Wakefield, Murch, Anthony, Linnell, Casson, Malik, Berelowitz, Dhillon, Thomson, Harvey, Valentine, Davies and Walker-Smith’s (1998) proposed a causal link between the MMR vaccine and autism and Crohn’s disease (a form of irritable bowel disease: IBD), vaccine uptake rates gradually declined in the UK. Parents of young children began to question the safety of MMR immunisation and were required to assess the risks and benefits of MMR immunisation during their decision-making process. The studies in the present thesis aimed to investigate factors influencing parents’ risk assessment, MMR intentions and behaviour to gain an understanding of parents’ decision-making process. A mixed method approach was taken, using both quantitative and qualitative methods. Four studies involving parents of young children and students were carried out in Scotland. The first study was a cross-sectional questionnaire-based study which used Protection Motivation Theory (PMT) and Subjective Norm (SN) to understand parents’ MMR immunisation intentions and behaviour for first dose and second dose MMR vaccine. The results suggest PMT was a useful psychometric risk model when examining first-dose and second-dose MMR immunisation and associated risks. The inclusion of SN in the model increased its overall robustness. Differences between immunising parents and non-immunising parents were identified. Immunisers perceived measles, mumps and rubella to be severe diseases and reported greater susceptibility and fear in relation to the diseases, whereas non-immunisers reported more concern about the associated risks of autism and IBD. Additionally, immunisers were more likely to follow the advice of health professionals (GP and health visitor) and reported them to be important sources of information, whereas non-immunisers were less likely to follow advice from health professionals and reported the media and internet to be important sources of information. The second study used a similar methodology to the first study but used PMT and SN to investigate MMR immunisation decision-making in a sample of 90 previously non-immunised university students during a mumps outbreak on campus. PMT and SN were found to be important constructs when understanding the students’ immunisation behaviour. In comparison with non-immunising students, immunising students reported greater fear, severity and perceived risks of the vaccine-preventable diseases. Non-immunisers perceived greater external barriers to immunisation and anxiety about immunisation to be an internal barrier. Both groups valued the information provided by health professionals and were more likely to follow the advice from these referents than any other referent group. Comparisons were made between the results of the first two studies. The findings indicate PMT, including SN, was a useful model when examining the MMR decision-making process for immunisers and non-immunisers in different population groups. Parents and students reported similar threat appraisals in relation to the vaccine-preventable diseases, but were different in their coping responses (response efficacy and self-efficacy). Many similar patterns between students and parents were illustrated, but parents reported stronger beliefs related to their parental role. The results indicate that MMR decision-making differs depending on the population under study. The third study used 5 focus groups and thematic analysis to explore the role of subjective norm (SN) and other social norms in greater depth with immunising parents. The findings indicate that social norms play a central role in the decision-making process, in addition to SN. Social normative factors which were found to contribute to the decision process included: group identification and norms, SN, descriptive norms, private self, relational self, and moral norms. The ‘private self’, i.e. own personal identity as ‘parents’, and feelings of moral obligation to their child were perceived as important social norms during the MMR decision. Parents were willing to listen to the advice of significant others but perceived their ‘private self’ as playing a more active role during the decision process. Experience of other parents’ MMR behaviour (descriptive norm) contributed to the risk assessment of the MMR vaccine and increased confidence in their own decision when congruent. The dual role of health professionals (who were also parents) as a ‘medical professional’ and as a ‘parent’ was influential during the decision process. The final study further investigated the influence of health professionals (HPs) on parents’ MMR decision-making. The role of interpersonal and generalised trust was explored using one-to-one interviews with 6 MMR immunisers, 3 non-immunisers and 8 immunisers with single vaccines. All groups of parents reported generalised mistrust in the Government based on the provision of biased information and past experiences of Government behaviour. Parents who opted for the MMR vaccines described interpersonal trust with their own HPs, where HPs were willing to openly discuss concerns relating to the MMR vaccine. Parents opting for the single vaccines or refusing all vaccines tended to report mixed experiences with their HPs, with some parents citing them as unhelpful and unwilling to discuss MMR concerns. Greater trust was illustrated by all parents, regardless of immunisation status, for their own health professionals and the NHS than for private clinics offering the single vaccines. Parents opting for the single vaccines perceived them to be safer (in terms of autism and IBD) than the MMR vaccine but questioned their credibility. The four studies illustrate that PMT facilitates understanding of parents’ MMR decision making and behaviour, and highlights the importance of including social norms (as well as important sources of information) and trust in future MMR immunisation research. Furthermore, comparisons with parents and students illustrate differences in coping appraisal between the two groups and suggest risk assessment differs depending on the saliency of the risk for the population group.