tb [25], [26], [29] and [30]. The same pattern was seen for this cytokine, such that immunisation with 50 μl induced a greater number

of antigen-specific CD8+IL17+ cells in the lung than immunisation with 5–6 μl. The presence in the lung of antigen-specific CD8+ T-cells of an effector phenotype, defined by the level of expression of CD62L and CD127 [22], correlates with protection after Ad85A immunisation [9]. Here we show that immunisation with Ad85A in 50 μl i.n. induces a significantly greater number of antigen-specific effector and effector memory cells in the lung than immunisation in 5–6 μl (Table 2). These phenotypic data indicate that immunisation with 50 μl generates a consistently greater number of antigen-specific CD8+ T-cells TGF-beta family in the lung than 5–6 μl, whether these cells are detected by production of IFNγ, IL2, TNF or IL-17, suggesting that the number of 85A-specific CD8+ T-cells in the lung at the time of challenge is the most important factor determining TSA HDAC the extent of protection against M.tb. We suggest that i.n. immunisation with 50 μl Ad85A has two important effects. The first is that antigen delivered to the deep lung [18] induces an immune response in the draining mediastinal nodes, and the second is that the adenovirus induces inflammation in the lung. This means

that antigen-specific cells leaving the mediastinal lymph nodes and passing via the thoracic duct, the right side of either the heart and pulmonary

arteries, will be recruited back to the lungs, including the airways, because of local inflammation [31]. Any activated, non-antigen-specific cells in the blood will most likely also be recruited into the lungs. In contrast, immunisation with a small volume induces a weak immune response in the NALT and perhaps the cervical nodes, but because the lungs are not inflamed, cells leaving these inductive sites will not be preferentially recruited to the lungs. Additionally, because the mechanisms of homing are partially shared between different mucosal tissues, it is possible that cells induced in the NALT might return to the bronchial-associated-lymphoid-tissue (BALT) or to the mucosa of the large airways of the lung [12]. This may provide another explanation why NALT-induced cells provide little or no protection, as it is the presence of cells in the airway (bronchioles and alveoli) that has been correlated with protection [7] and [8]. Alternatively, since it is known that mucosal responses are sometimes tolerising, it may be that in the absence of a mucosal adjuvant the NALT environment generates non-protective T-cells [32]. The importance of targeting both respiratory and other mucosal pathogens at their site of entry is becoming more apparent.