However, this phenomenon did not occur when executing the experiment with the PB28 dihydrochloride serovar Lai counterparts. This serovar displayed similar responses irrespective of its killed or viable status and for both hostadapted and culture-adapted strains. There are two possible explanations for this observation. Firstly, differences in responses might reflect differences in expression of potential virulence factors in the two serovars. The host-adapted serovar Lai showed a markedly lower expression of LigA compared to the host-adapted serovar Bataviae, which is consistent with its higher number of in vitro passages. Differential expression might very well have occurred for other bacterial cell components. In the guinea pig model, both the host-adapted serovar Lai and Bataviae escaped complement Papaverine hydrochloride killing in the blood, enabling Lai to adapt, grow and disseminate, thus regaining its full host-adapted constitution. However, incubation time and conditions with THP-1 cells most probably did not enable such adaptation and results therefore might reflect a host interaction with a pathogen sharing more features similar to a culture-adapted than to a host-adapted form. This option is important as it suggests that virulence testing of Leptospira strains in animals alone may not be a reliable marker for their in vitro performance. Secondly, it cannot be excluded at this stage that serovar specificity does account for the differences in cytokine responses by THP-1 cells. In this respect it should be noted that, whereas no specific clinical manifestations can be associated with distinct serovars, some serovars tend to cause more serious disease than others. Both explanations for different in vitro behaviour of the two host-adapted strains are highly relevant as these will have consequences for future interpretations of in vitro experiments on innate immune responses. Therefore, it is important to further investigate the role of differential gene expression and serovar specificity in innate immune responses. Differences in the outcome of in vitro experiments apparently is also associated with the type of host cells used in the tests because different TNF-a release profiles were observed when using PBMC compared to THP-1 cells.