Paratuberculosis is a chronic progressive granulomatous enteritis of ruminants caused by Mycobacterium avium subsp. paratuberculosis (MAP). Paratuberculosis in cattle is clinically characterized by weight loss, emaciation and diarrhea and subclinically by reduced milk production leading to considerable economic losses to farming community. Paratuberculosis is a staged infection in which young calves acquire the infection in the first months of life, may progress into a prolonged asymptomatic stage of about 2-5 years and may eventually become clinically infected animals. Vaccination with whole-cell live or inactivated vaccines prevents or delays the development of clinical stage of the disease but does not eliminate MAP and is usually accompanied by interference with bovine tuberculosis diagnostics as well as local tissue damage. Subunit vaccines with well-defined antigens in combination with a suitable adjuvant offer the possibility to avoid these limitations and induce a strong T helper 1 (TH1) type immune response that has been associated with protection against MAP. The aim of the study was to identify proteins from different stages of infection and formulate them into a multi-stage subunit vaccine with activation of protective immune response in experimentally challenged calves, with a focus on cell-mediated immune responses chiefly interferon gamma (IFN-γ) and polyfunctional T cells. The antigen composition of the vaccines was selected based on previous immunogenicity studies in cattle and experimental knowledge from in vitro and in vivo expression studies with M. tuberculosis proteins in mice (101). The vaccines were used to investigate the influence of age on vaccine-induced T cell responses and measuring vaccine-induced protective efficacy after experimental challenge. Effect of costimulation on vaccine-induced T cell responses and immune correlates of vaccine-induced protection were further characterized. Early expressed and latency-associated MAP proteins were formulated in cationic adjuvant formulation (CAF) 01 and tested in calves through two different experiments, MAP multi-stage vaccine (MSV)-1 and 2. FET11 vaccine, a combination of a fusion protein of four early expressed MAP proteins and a latencyassociated MAP protein formulated in CAF01 adjuvant was tested by an experimental MAP challenge in calves. FET11 vaccination at 16 weeks of age induced significant immune response and conferred protective immunity characterized by a mean 1.1 log10 reduction in bacterial numbers in the gut tissues compared to control animals and was superior to a commercial whole-cell heat inactivated vaccine, Silirum® or FET11 vaccination at 2 weeks (Article 4). In both MSV experiments, most significant immune responses were observed against Esx-secretion system proteins and latency proteins (Article 1 and 4). However, the immunogenicity of two recombinant MAP proteins common in both studies was different, emphasizing the possibility of dynamics of MAP infection guiding the differential immune response. There was an association between age of vaccination and induced immune responses. Older animals (4 months) developed a more robust immune response (Article 1 and 4). Furthermore, no significant increase in the immune response was observed 8 weeks after second booster vaccination in MAP MSV-1 study (Article 1). However, decreased immune responses after one year period in MAP MSV-2 experiment, warrants the use of a booster vaccination. The experimental challenge of calves with midlog-phase frozen stock MAP cultures correlated well with whole-blood IFN-γ responses to PPDj in advanced weeks of the study, which signifies PPDj response as a marker of MAP experimental infection (Article 4). This challenge study also supports the possibility of establishing a uniform and repeatable bovine MAP infection model involving large number of animals procured at different times. The results also show the potential application of quantitative real-time PCR (qPCR) for the evaluation of microbial load in tissues and vaccine efficacy (Article 4). Costimulation of vaccine-induced ex vivo T cells significantly increased IFN-γ levels following use of anti-CD28 and anti-CD49d antibodies (Article 2). Recombinant interleukin IL-12 (rIL-12) also resulted in very high levels of IFN-γ production but was accompanied by high background levels. Thus enhanced antigenspecific immune response with anti-CD28/CD49d costimulation could be suitable for characterizing vaccination or infection-mediated responses, while rIL-12 with a more TH1 biased potentiation of antigen-specific IFN-γ production warrants its use for diagnostic purposes. The results of this study also highlight the role of innate immune cells such as gamma delta (γδ) T cells and natural killer (NK) cells in paratuberculosis infection (Article 2). Antigen-specific IFN-γ production by γδ T cells and NK cells was observed in vaccinated calves. Although the levels were low compared to CD4+ T cells, IFN-γ production by these innate effector cells might compensate the immature immune system of young calves to counteract MAP infection. A number of immunological markers were discussed as potential vaccine-induced immune correlates with emphasis on infections requiring T cell-mediated immunity (Article3). Traditionally, neutralizing antibody titers are associated with vaccine-mediated immune protection. However, advancement of biological techniques has allowed identifying and appreciating immune markers as candidates for novel correlates of protection. Taken together, this study has provided information on developing a multi-stage vaccine against paratuberculosis and has increased the knowledge regarding age of vaccination, experimental MAP infection, costimulation signals for measuring T cell responses, and immune correlates of protection.