Etude de Coxiella burnetii, agent de la fièvre Q, en Belgique : prévalence, typage moléculaire et recherche de nouveaux outils de diagnostic et de contrôle

Q fever is a worldwide zoonosis caused by an intracellular bacterium, Coxiella burnetii. Small ruminants (goats and sheep) are considered as the most common animal reservoir for human infection. Since the recent Q-fever Dutch outbreak, the initial objective of this PhD thesis was to study the situation of this disease in Belgium. Q fever is endemic in our country. It is present in goat (herd prevalence ranging from 6.3 to 12.1%) and bovine (etiologic agent of 1% of abortions) herds. Monitoring of goat bulk tank milks and bovine abortion products allowed genotyping of C. burnetii strains circulating in our country. The genetic profiles of the caprine samples were of a great variability than the bovine samples which displayed a more homogeneous pattern. The reference Nine Mile strain (isolated from ticks in the United States) had a genetic profile different from all the other isolates. These initial results prompted us to hypothesize the presence of host specificity within the C.burnetii species. Totally, five caprine (CbBEC1-5) and one bovine (CbBEB1) strains were isolated. In milk, two shedding profiles were identified which were either continuous (type I) and associated to the Dutch genotype; or discontinuous (type II) and linked to other genotypes. In parallel with field studies, we developed experimental models, in vitro (cell lines) and in vivo (mice). With these models, we showed that there was no hyper-virulent behavior displayed by the Dutch-type caprine strain and we reinforced the host specificity hypothesis previously formulated. Indeed, the strain isolated from cattle showed greater proliferative potential than the caprine strain in allogeneic bovine macrophagic cell line. Another objective of this work aimed at improving the current diagnostic tools and vaccination. By phage display analysis, we looked for peptides mimicking the lipopolysaccharide (LPS) of C. burnetii. A monoclonal antibody directed against the virenose (specific sugar of the C. burnetii LPS) was used for selection. Two phage-enzymes were identified. Tests to confirm their diagnostic application have failed because of low specific binding. The last part of the study dealt with vaccination, so as currently applied in field setting. In Belgium, goat positive farms are vaccinated with Coxevac®, the only authorized vaccine. This vaccine is produced with a strain not circulating in Europe (Nine Mile). Use of Coxevac® led to a reduction in shedding of C.burnetii but did not prevent infection of new farms. Moreover, vaccination reduced the excretion of bacteria after a period of 5 to 6 months and several injections were needed. Given the weaknesses of the current vaccination program, we tested new vaccine candidates in the mouse model. Vaccines prepared from CbBEC1 and CbBEB1 strains showed better protection than the Nine Mile-derived vaccine. These results further supported the hypothesis of host specificity within C.burnetii. The protective efficacy of the two mimetic peptides selected by phage display was not relevant. The tools developed along with this PhD work were of paramount importance 1) to improve the expertise of the animal Q fever National Reference Laboratory, 2) to conduct a deep epidemiological study in the field, and 3) at the research level, to better understand the dynamic of infection of C.burnetii in experimental models.

(fre) La fièvre Q est une zoonose de répartition mondiale qui est due à une bactérie intracellulaire : Coxiella burnetii. Les sources principales de contamination pour l’homme sont les produits d’avortement des petits ruminants domestiques. Suite à l’épidémie des Pays-Bas, le premier objectif de ce travail était d’étudier la situation de la fièvre Q en Belgique. Nous avons déterminé que la fièvre Q est endémique dans notre pays. Elle est présente dans le cheptel caprin (6.3 à 12.1% de prévalence troupeau) et bovin (agent étiologique de 1% des avortements). L’observation d’un regroupement génotypique en fonction de l’espèce hôte, nous a amené à émettre l’hypothèse qu’au sein de l’espèce C. burnetii, il pourrait y avoir des sous-types différents (hypothèse de spécificité d’hôte). Ce suivi nous a également permis de déterminer deux types d’excrétion dans les laits de tank caprins. Au vu de la diversité des génotypes, nous avons mis au point, en parallèle de l’étude sur le terrain, des modèles d’infection in vitro (lignées cellulaires) et in vivo (souris). Dans ce travail, nous avons également essayé d’améliorer les outils de diagnostic et de vaccination en recherchant des mimes peptidiques du lipopolysaccharide de C. burnetii par phage display. L’amélioration de l’outil vaccinal suivait deux axes. Nous avons, d’une part, analysé la vaccination mise en place par les autorités belges. D’autre part, étant donné les faiblesses du vaccin actuel, nous avons testé de nouveaux candidats vaccins dans un modèle vaccinal murin.