UC Berkeley

Listeria monocytogenes is a Gram-positive bacterial pathogen that has been used for decades to study bacterial infection and immunity. Murine infection models demonstrate that sublethal infections with L. monocytogenes cause only a mild disease but lead to the development of long-lived cell-mediated immunity (CMI), mediated by antigen-specific cytotoxic CD8+ T-cells. During infection L. monocytogenes is phagocytosed by antigen presenting cells but escapes the phagosome and accesses the host cell cytosol by secreting the virulence factor LLO, encoded by the gene hly. Importantly, robust immunization requires live bacteria that can enter the cytosol of host cells. In contrast, Δhly mutants remain in the phagosome (do not enter the cytosol) and are poor inducers of CMI. Furthermore, during co-administration of both cytosolic and Δhly strains simultaneously into mice, the Δhly mutant blocked CMI, indicating that the Δhly mutant suppresses immunity normally elicited by the cytosolic strain. Additionally, Δhly mutants induce the expression of the immunosuppressive cytokine IL-10, that inhibits the expression of inflammatory cytokines and the development of CMI. In chapter two, we identify the genetic and molecular determinants of IL-10 induction and immunosuppression by phagosome confined bacteria by screening 6000 Δhly L. monocytogenes transposon mutants for either enhanced or diminished capacity to induce IL-10 in macrophages. The transposon screen identified two classes of bacterial mutants. The first class led to diminished IL-10 induction and contained a transposon insertion in the gene that encodes the enzyme Lgt that is required for the formation of mature bacterial lipoproteins. We found that macrophages induced elevated IL-10 secretion in response to phagosome confined bacteria by detecting bacterial lipoproteins through the host receptor TLR2. The second class of mutants led to greater IL-10 induction and contained transposon insertions in the genes encoding two known cell-wall modifying enzymes, pgdA and oatA. The ΔhlyΔpgdAΔoatA triple mutant was known to be susceptible to lysozyme-dependent killing within the macrophage and induced over 200% more IL-10 than the Δhly mutant. This suggested that killed phagosomal bacteria were being degraded and thus releasing their nucleic acids within the phagosome that were then being detected by nucleic acid sensing phagosomal TLRs, leading to the robust induction of IL-10. Macrophages lacking both phagosomal TLRs and TLR2, no longer secreted IL-10 during infection with the ΔhlyΔpgdAΔoatA mutant. Although both TLR2 and endosomal TLR pathways led to IL-10 induction in macrophages, the detection of nucleic acids was the main immunosuppressive pathway since mice defective for phagosomal TLR localization were no longer suppressed for the development of CMI by the Δhly mutant. In conclusion, the work in chapter two identified both the bacterial and host molecular determinants that lead to immunosuppression during vaccinations with phagosome confined L. monocytogenes Δhly mutants. The work presented in chapter three sought to enhance L. monocytogenes-based cancer vaccines. A current effort in the field of tumor therapy involves converting tumors that respond poorly to cancer immunotherapy, termed immunologically “cold” tumors, into “hot” tumors by injecting antitumoral agents directly into the tumor microenvironment. However, the mechanisms that lead to effective intratumoral therapy are not fully understood and it is not clear whether it is beneficial to target the tumor cells or the immune cells during intratumoral therapy. The work in chapter three demonstrates that intratumoral injections with cytosolic L. monocytogenes strains are therapeutic in mice, whereas phagosome confined mutants are poorly therapeutic. In vitro infections demonstrated that L. monocytogenes infected macrophages approximately 100 times more efficiently than tumor cells. Nevertheless, in vitro infection of tumor cells suggested that L. monocytogenes could kill tumor cells directly by overgrowing within the tumor cell cytosol. We, therefore, hypothesized that enhancing L. monocytogenes’ ability to invade tumor cells would also enhance intratumoral therapy by enhancing the cold to hot conversion of tumors. A selection was performed for spontaneous L. monocytogenes mutants that were enhanced for invasion of tumor cells in vitro, with the goal of identifying the major bacterial determinants that contribute to tumor cell invasion, and to determine if invasion of tumor cells contributes to I.T. therapy. The selection identified two gain of function mutations in the master virulence regulator PrfA, that lock PrfA into a constitutively active form, termed PrfA*. Additionally, a novel mutation in the virulence gene ActA (ActAE144L) was identified that also conferred enhanced invasion into mouse tumor cells, indicating that ActA could be used to test if enhanced invasion contributes to intratumoral therapy. However, ActA did not play a beneficial role during intratumoral therapy in mice. Analysis of L. monocytogenes mutants identified that the majority of PrfA*-dependent-enhanced invasion was due to the virulence factor InlB, an internalization factor known to facilitate entry into non-phagocytic hepatocytes that express the hepatocyte growth factor receptor c-Met. Finally, InlB contributed significantly to intratumoral therapy in two commonly studied types of mouse tumors, CT26 and B16.F10. In conclusion, this work showed that in vitro genetic selections with attenuated bacterial pathogens can identify the relevant genes and pathways that affect tumor cell invasion. These results revealed that bacterial internalization factors that affect the invasive capacity of L. monocytogenes into non-phagocytic cells contributed significantly to intratumoral therapy.