S2, A and B). poly(I:C)-mediated NK cell activation was not notably impaired in mice lacking CD8 DCs, providing further evidence that poly(I:C) functions through diverse accessory cells rather than solely through DCs. These results demonstrate distinct yet complementary Estropipate functions for MDA5 and TLR3 in poly(I:C)-mediated NK cell activation. Microbial components play a major role in activating innate and adaptive immune Itgb5 responses by triggering pattern acknowledgement receptors (Ishii et al., 2008). Poly(I:C) is an analogue of viral double-stranded RNA (dsRNA) that activates numerous immune cell types through two major dsRNA sensors, melanoma differentiation-associated protein-5 (MDA5) and Toll-like receptor 3 (TLR3). MDA5 is usually a cytosolic sensor, which Estropipate detects poly(I:C) that penetrates into the cytosol through as yet undefined mechanisms (Ishii et al., 2008). TLR3 is located in intracellular endosomes and detects poly(I:C) that has been internalized by endocytosis (Matsumoto and Seya, 2008). Upon poly(I:C) detection, MDA5 transmits signals through the adaptor IPS1, whereas TLR3 signals through the adaptor TRIF (also known as TICAM1). Both of these adaptors initiate downstream signaling pathways that lead to activation of a similar array of transcription factors, including IRF3, IRF7, IRF1, and NF-B. These factors induce the expression of genes encoding type I IFNs: i.e., IFN- and IFN-, proinflammatory cytokines, and various molecules involved in antigen presentation (Kawai and Akira, 2008). Poly(I:C) induces the maturation of DCs, improving their ability to Estropipate primary and expand antigen-specific T cell responses (Kumar et al., 2008;Trumpfheller et al., 2008;Longhi et al., 2009). Because of this DC stimulatory activity, poly(I:C) is usually a promising adjuvant for vaccines, particularly for malignancy vaccines that must overcome both tolerance to tumor-associated self-antigens and the immunosuppressive influence of the tumor microenvironment (Steinman and Banchereau, 2007). Poly(I:C) is also extensively used to activate mouse NK cells in vivo. The NK cell stimulatory activity of poly(I:C) is usually potentially important for anticancer vaccines, as it may contribute to tumor Estropipate eradication by inducing NK cellmediated lysis of tumor cells. In humans, the NK cell stimulatory activity of poly(I:C) has been chiefly attributed to its ability to trigger TLR3 expressed in cultured NK cells (Schmidt et al., 2004;Sivori et al., 2004;Hart et al., 2005;Lauzon et al., 2006). Whether poly(I:C) also activates human NK cells through DCs or other accessory cells has not been investigated. In mice, poly(I:C) is usually thought to activate NK cells primarily through DCs. An initial study showed that poly(I:C) stimulates TLR3 in DCs, which consequently acquire the ability to activate NK cells (Akazawa et al., 2007). A very recent study showed that poly(I:C) triggers both the TRIF and IPS1 signaling pathways in CD8 DCs, which in turn activate NK cells in vitro (Miyake et al., 2009). These results suggest that TLR3 and MDA5 may stimulate mouse NK cells indirectly through activation of DCs, particularly CD8 DCs. In this study, we investigate the relative contributions of MDA5 and TLR3 in poly(I:C)-mediated activation of NK cells using MDA5/, TLR3/, and MDA5/TLR3/mice. We find that MDA5 has a predominant role in NK cell activation, whereas the contribution of TLR3 is usually secondary and is most obvious in the absence of MDA5. Both MDA5 and TLR3 activated NK cells indirectly through accessory cells but induced different NK cell stimulatory cytokines, as MDA5 was essential for IFN- and IFN-, whereas TLR3 was required for IL-12 and, in part, for IFN-. By generating BM chimeras between WT and dsRNA sensor-deficient mice, we found that MDA5 promotes NK cell activation mainly through stromal accessory cells, whereas TLR3 functions predominantly through BM-derived accessory cells. To determine whether the hematopoietic accessory cells that activate NK cells are indeed CD8 DCs, we examined mice deficient for the basic leucine zipper transcription factor ATF-like 3 (BATF3), which have a selective developmental defect in CD8 DCs.