models of live Mycobacterium tuberculosis infection (11). Whether vaccines containing killed M. tuberculosis would also induce MDSCs remains unclear.Among the carrier substances for antitumor or antipathogen vaccines, incomplete Freund's adjuvant (IFA), used under the name Montanide ISA-51 VG, is a suitable and widely distributed mineral oil-based adjuvant for human clinical studies with good antigen responses but also some adverse effects (12,13). Another preparation, CFA, was developed in 1942 and since then has been widely used as a vaccine adjuvant in experimental animals (14). Since CFA is composed of IFA containing heat-killed and dried M. tuberculosis, it can be considered as an M. tuberculosis vaccine. Notably, unclear tolerogenic effects had been observed after CFA vaccination in animals (14). On the one hand, CFA is a critical component for the induction of autoimmunity in models of experimental autoimmune encephalomyelitis (15, 16); on the other hand, injection of CFA could also prevent autoimmunity, such as the onset of type I diabetes in NOD mice (17) and ameliorated Parkinson's disease in rats (18). While the pathogen-induced adjuvant effects can be explained by mycobacterial triggering of TLR2 and TLR4, the tolerogenic mechanisms of CFA are not understood. The possible suppressive effects, besides their immunostimulatory function by IFA and other adjuvants in tumor vaccination studies, have been reviewed before (13). CFA-based tumor vaccines could induce suppressive CD11b + immune cells (7,19), while in an alum-based vaccine, B cell immune responses were boostered by CD11b + cells (19). However, these questions remain unclear: whether M. tuberculosis or IFA/Montanide will be able to induce MDSCs, which MDSC subsets will become activated, and what are the major target cells of their suppression in vivo.Previously, we described a simple and reliable protocol to generate murine MDSCs in vitro from murine BM precursor cells with . The generation of such human or murine M-MDSCs in vitro can be performed by following a 2-step protocol that, first, converts classical monocytes (Ly6C hi or CD14 + ) into monocytes "licensed" for suppression (L-Mono) that also can be considered as resting M-MDSCs and, second, converts L-Mono into activated M-MDSCs that release the suppressor molecules NO in the murine and IDO in the human system (21). This in vitro differentiation model exemplifies the 2 steps of signaling for M-MDSC induction that are relevant also in vivo (1). While the combination of LPS and IFN-γ for the second step of MDSC activation constituted one of the strongest signals (22), cocktails of proinflammatory cytokines were also effective (21). While the infiltration of MDSCs at vaccination sites and lymph nodes has been observed before using Mycobacterium bovis BCG, only local T cell responses seemed to be affected in these organs (23).The targets of MDSC-mediated suppression were mainly identified as macrophages, NK cells, and CD4 + or CD8 + T cells that were tested for their phagocytic activity, proli...