Cytokine Gene and Protein Expression in BCG Vaccinated and Non-Vaccinated Mycobacterium bovis Infected cattle
The persistent increase of bovine tuberculosis (bTB) over the past twenty years has put a substantial strain on both the British economy and the welfare of livestock. However, the development of an effective bTB vaccine has been continually hindered by the lack of knowledge on the immune response following Mycobacterium bovis (M. bovis) infection. In collaboration with the TB Research Group at the Veterinary Laboratories Agency (VLA, Surrey), this thesis is part of a much wider strategy managed by the Department of Environment, Food and Rural Agency (DEFRA) aimed at elucidating the immunopathogenesis of M. bovis and to develop more effective infection control measures. The specific focus of this thesis was to enable a stronger understanding of the bovine immune response over different periods of M. bovis infection and to apply this new knowledge in evaluating the efficacy of a novel BCG vaccination. Time Course Study: Knowledge of time dependent cytokine expression following M. bovis infection would aid vaccine development by revealing potential correlates of protection. Interferon gamma (IFN-γ), tumour necrosis factor alpha (TNF-α), interleukin (IL) 4 and 10 expression were analysed using quantitative (q) PCR in formalin fixed bovine lymph nodes following five, twelve and nineteen weeks of M. bovis infection. A strong pro-inflammatory/ T helper 1 (TH1) lymphocyte response was evident at five weeks post M. bovis infection, represented by IFN-γ and TNF-α expression (log2 copies of 6.5 and 2.15, respectively) in the absence of IL4. Between five and twelve weeks of infection, a significant increase was observed in IL10 (log2 copies from 5.97 to 8.27, p<0.01, Mann Whitney test), accompanied by an increase in both IFN-γ (log2 7.53) and TNF-α (log2 3.94). This data conformed to a recently described aspect of TH1 lymphocytes, a ‘self-limiting’ nature in which cells produced both IFN-γ and IL10 with the aim of controlling the heightened pro-inflammatory response. The role of IL10 as an immunosuppressive became evident when comparing cytokine expression between four different types of thoracic lymph node; the left bronchial (LB), cranial mediastinal (CRM), caudal mediastinal (CM) and cranial tracheobronchial (CRT) nodes. The LB and CRM lymph nodes produced significantly higher levels of IFN-γ expression (log2 copies between 8.2 and 10) as compared to the CM and CRT (log2 copies between 2.6 and 5.5, p<0.001, Mann Whitney test). Further analysis of the data as a profile of cytokine expression for each lymph node type revealed that IFN-γ was dominantly expressed within the LB and CRM nodes, whereas within the CM and CRT nodes, IL10 was the dominant cytokine. The former nodes also displayed a higher level of pathological damage (represented by mean percentage area coverage of granuloma, 33.6 and 20%, respectively) as compared to the CM (13%) and the CRT lymph node types (10.8 %). This suggests conflicting roles for IFN-γ and IL10 in the development of immune-associated pathology. Following nineteen weeks of infection, the expression levels of IFN-γ, TNF-α and IL10 reduced (log2 6.22, 3.02 and 7.03, respectively) implying a loss of the cellular response. The later stages of bovine tuberculosis have been shown within the literature to display characteristics of a humoral rather than cell mediated response. However, within this study at nineteen weeks post infection IL4 (an important cytokine in the development of the humoral response) remained undetectable. The results from this study therefore confirm the importance of the cell mediated immune profile in response to M. bovis infection as well as the integral role of IFN-γ in both protection and pathology. It also further demonstrates the involvement of IL10 in controlling the IFN-γ response and highlights this cytokine as being potentially important in future immunologybased vaccination studies. BCG Vaccination Study: The current vaccine used against human tuberculosis, BCG, has provided variable results on protection against infection in experimental bovine studies. The BCG bacterium has lost a comparatively large quantity of genomic DNA through attenuation since its primary production in 1921, of which the majority represented genes encoding antigenic proteins. MPB70 and MPB83 are differentially expressed between BCG sub-strains due to a single nucleotide polymorphism in the alternative sigma factor K (SigK). BCG Pasteur has been shown to produce low levels of these antigenic proteins; however complementation of BCG Pasteur with a copy of sigK from BCG Russia resulted in up-regulating expression. It was therefore hypothesised that the recombinant BCG (sigK) Pasteur would prove more efficient in controlling M. bovis infection by inducing a stronger protective immune response post vaccination. IFN-γ, TNF-α, IL 4 and 10 expression were analysed using qPCR within the freshly dissected lymph nodes of five experimental cattle groups; BCG Pasteur vaccinated M. bovis challenged, BCG (sigK) Pasteur vaccinated challenged, non-vaccinated infected, non-vaccinated noninfected and BCG Pasteur vaccinated non-infected. Five weeks following infection, a strong IFN-γ mRNA response was detected in both the non-vaccinated and vaccinated cattle (mean log2 copies between 9.6 and 10.5 as compared to between 7.84 and 8.58 in the non-infected cattle). M. bovis infection also induced a significant reduction in IL10 mRNA levels in both vaccinated and non-vaccinated cattle (mean log2 14.4 in the infected groups compared to 15.5 in the non-infected cattle, p<0.005, Mann Whitney test) although there was little difference in TNF-α expression (mean log2 copies between 11.06 and 11.8 in all five groups). Interestingly, IL4 mRNA was detectable only within the two non-infected control groups (mean log2 12.4), further supporting the concept of a strong cell mediated response after five weeks of infection. Vaccination prior to challenge had an effect on IFN-γ mRNA levels only, as both the BCG Pasteur and BCG (sigK) Pasteur vaccinated groups displayed a smaller increase in IFN-γ mRNA following challenge (mean log2 10.3 and 9.6, respectively) as compared to the nonvaccinated group (mean log2 10.5). This reflected the role of vaccination in priming the immune response to enable more rapid elimination of the bacteria and subsequently inducing a lesser pro-inflammatory response. Interestingly, the BCG Pasteur vaccinated group appeared to control the immune response to a greater extent, as IFN-γ mRNA was significantly similar to that observed in the non-vaccinated non-infected group (mean log2 8.58, p>0.05, Mann Whitney test). In addition to the qPCR data, levels of IFN-γ and TNF-α protein (represented by the number of cells producing these proteins) were also analysed by immunohistochemistry. IFN-γ protein in the five experimental groups displayed the same pattern as that observed for IFN-γ mRNA expression (p<0.001, Spearmans correlation coefficient). However, analysis of TNF-α protein revealed significant differences between the five groups (p<0.005, Kruskal Wallis test) in contrast to that observed for the mRNA levels (p>0.05, Spearmans correlation coefficient) suggesting that posttranscriptional controls may play an important role in TNF-α translation. The difference in IFN-γ mRNA and protein expression between the two vaccination groups was also reflected within the pathological data. Although both BCGs reduced levels to below that of the non-vaccinated group (represented by mean percentage area coverage of granuloma, 59%), the BCG Pasteur group displayed less pathology (mean 6%) compared to the BCG (sigK) Pasteur cattle (mean 35%). It was suggested that the increased antigenic repertoire of the recombinant BCG (sigK) Pasteur did result in a stronger stimulation of the immune response post vaccination but that, as a consequence the bacterial threat was eliminated more rapidly. This resulted in shortening the duration of antigenic stimulation thereby effecting the development of the memory T cell response. These results imply that enhancing the antigen repertoire of the current BCG alone is not sufficient in improving upon protection against M. bovis infection. They further support the benefits of a prime/boost vaccination protocol, in which primary antigenic stimulation of the bovine immune response is boosted at a later stage.