Despite this, the lactobacilli inhibited IL-13 induction, regardless of donor, either allergic or not. In the long-term cultures and the αCD3/αCD28-stimulated cultures, the increased IFN-γ and IL-12 Y-27632 cell line secretion in hPBMC cultures exposed to the lactobacilli could mediate the Th2-suppressive effect, as observed previously (Pochard et al., 2002; Bickert et al., 2009). However, the Th2 cytokine inhibition could be dependent on several parameters depending on the strains used (Pochard et al., 2002; de Roock et al., 2010; Lopez et

al., 2010). The exact mechanism by which probiotic lactic acid bacteria modulate the host immune response is largely unknown. Bacterial cell surface macromolecules (such as long surface appendages, extracellular polysaccharides and teichoic acids) are in direct contact with several immune cell types by binding various pattern GSK2126458 recognition receptors of the host. The structure

of the main cell wall macromolecules is strongly conserved, but various modifications, such as glycosylation and also quantitative differences, can contribute to the strain-specific properties of probiotics. As little information is available regarding the specific bacterial components that for example induce the expression and production of cytokines, advances can be made in this area through the sequencing of genomes and transcriptomes that can be correlated to measured effects and enable testing which bacterial genes and stiripentol derived components are essential to specific immunomodulatory properties (Borchers et al., 2009; Fink, 2010; Kleerebezem et al., 2010; Lebeer et al., 2010; Meijerink et al., 2010). Large numbers of candidate strains are often tested as probiotics for immunomodulating properties in a variety of in vitro models

to select those strains with the best characteristics. In these in vitro studies, effects of heat-killed bacteria may not be directly extrapolated to effects of viable bacteria. Nevertheless, recent literature shows similar effects comparing live bacteria with heat-killed bacteria or even with components from the respective bacteria (Ghadimi et al., 2008; Li et al., 2009; van Hoffen et al., 2010). Very limited information is available with respect to the in vivo molecular responses to probiotic bacteria in human mucosal tissues; however, a recent study of van Baarlen et al. (2009) showed a considerable overlap between in vivo human responses to live and heat-killed L. plantarum, provided that these bacteria were harvested from the same phase of growth. Systematic studies to link in vitro data to in vivo effects have rarely been performed so far and results are also not found to be consist (Foligne et al., 2007). Based on the limitations of the in vitro model, extrapolations to in vivo effects must therefore be considered with caution.