Toll-like receptors as mediators of of tumor cell death

Recently, it has been demonstrated that invasive but nonpathogenic Salmonella typhimurium i.t. injected into melanoma B16-bearing mice led to complete regression of the tumor, although Salmonella was unable to directly kill B16 tumor cells in vitro. Since we found high levels of IFN-γ in infected tumor masses in vivo, we wondered whether this cytokine might be involved in tumor cell death. In this work, we have demonstrated that IFN-γ shows a cytotoxic effect only when administered in combination with Salmonella. In addition, this combined treatment is able to enhance TLR transcription, mainly TLR2 and TLR3, in B16 cells, suggesting a possible link between tumor cell death and TLRs, as already proposed in some studies. Our hypothesis is strengthened by the finding that, unlike IFN-γ, IFN-α does not induce neither tumor cell death nor TLR transcription. B16 tumor cell death induced by either TLR2 ligand Pam3SCK4 or TLR3 ligand poly I:C treatment in combination with IFN-γ indicates that both TLR2-TLR1 heterodimer and TLR3 are involved in tumor cell death, although they likely activate independent mechanisms. On the contrary, neither TLR2-TLR6 nor TLR4 seems to be able to efficiently kill tumor cells neither alone nor in combination with IFN-γ. Notably, several evidences suggest that Salmonella in combination with IFN-γ activates TLR2 rather than TLR3. However, since we cannot definitely rule out the existence of a TLR3 ligand in Salmonella, further experiments using anti-TLR2 and anti-TLR3 blocking antibodies will be necessary. Furthermore, in this study we demonstrate that B16 tumor cells undergo a first necrotic event in the first 24 hours of combined stimulations inducing IL-6 and KC release, probably through NF-kB activation. Moreover, caspase activation analysis has demonstrated that B16 cells undergo apoptosis as well, but a later time. In this regard, we detected mainly caspase-7 cleavage after 48 hours of combined treatments. However, incubation with the necrosis inhibitor necrostatin-1 (nec-1) strongly reduces caspase-7 activation, suggesting that a necrotic event might be responsible for the subsequent apoptosis, that may be independent of TLR engagement. B16 dying cells release HMGB1 as well, a marker of necrotic cell death, that is able to enhance tumor immunogenicity. However, since Salmonella alone triggers HMGB1 release without inducing cell death, HMGB1 may be actively secreted in response to Salmonella, but may also be passively released by necrotic dying cells. Because of the important role of HMGB1 in tumor immunogenicity, future studies will be performed with the aim of understanding how B16 cells release HMGB1. Our results suggest that this may be achieved through a TLR4 MyD88-independent pathway. Interestingly, we demonstrate for the first time that poly I:C induces TLR3 translocation to B16 cell surface in combination with IFN-γ. In this way, poly I:C can improve TLR3-mediated response thus probably inducing a stronger caspase-7 activation with respect to Salmonella and TLR2 ligand. Finally, different experiments have highlighted the importance of IFN-γ in this tumor cell death mechanism. In particular, IFN-γ seems to sensitize tumor cells before receiving a subsequent stimulation, up-regulating TLRs involved in tumor cell death (e.g. TLR3) or through other unknown mechanisms. These results set the basis to improve the immunotherapy protocol developed in Maria Rescigno’s laboratory, combining IFN-γ i.t. injection with Salmonella i.t. injection or replacing Salmonella with Pam3CSK4, in order to obtain a more prompt therapeutic effect avoiding possible side effects of Salmonella. In addition, these results provide a rationale for introducing poly I:C treatment into electrochemotherapy approach, a new method applied to cutaneous cancers treatment, melanoma included. Since it is based on the permeabilization of the cell membrane by means of short and intense electric pulses, it would allow the entry of poly I:C into the cytosol of both tumor and immune cells. Therefore, TLR3 ligands may be involved in both the killing of tumor cells and in activating the immune system. Importantly, this new system may overcome the use of a systemic treatment with IFN-γ and could be used to treat melanoma avoiding possible side effects.