In vivo overexpression of Flt3 ligand expands and activates murine spleen natural killer dendritic cells.

Natural killer dendritic cells (NKDC) are a unique class of murine immune cells that possess the characteristics of both natural killer (NK) cells and dendritic cells (DC). Because NKDC are able to secrete IFN-gamma, directly lyse tumor cells, and present antigen to naïve T cells, they have immunotherapeutic potential. The relative paucity of NKDC, however, impedes their detailed study. We have found that in vivo, overexpression of the hematopoietic cytokine Flt3 ligand (Flt3L) expands NKDC in various organs from 2-18 fold. Flt3L expanded splenic NKDC retain the ability to lyse tumor cells and become considerably more potent at activating naïve allogeneic and antigen-specific T cells. Compared to normal splenic NKDC, Flt3L-expanded splenic NKDC have a more mature phenotype, a slightly increased ability to capture and process antigen, and a similar cytokine profile. In vivo, we found that Flt3L-expanded splenic NKDC are more effective than normal splenic NKDC in stimulating antigen-specific CD8 T cells. Additionally, we show that NKDC are able to cross-present antigen in vivo. The ability to expand NKDC in vivo using Flt3L will facilitate further analysis of their unique biology. Moreover, Flt3L-expanded NKDC may have enhanced immunotherapeutic potential, given their increased ability to stimulate T cells.

Adenovirus infection enhances dendritic cell immunostimulatory properties and induces natural killer and T-cell-mediated tumor protection.

Dendritic cells (DCs) are potent antigen-presenting cells with the potential for cancer immunotherapy. Adenoviral-mediated gene transfer is an attractive means to manipulate the immunostimulatory properties of DCs for therapeutic advantage. Because adenovirus induces DC maturation, we postulated that it would significantly alter their immune functions. Infected DCs markedly increased allogeneic and antigen-specific T-cell proliferation, and augmented natural killer cell lytic activity and IFN-gamma production. The enhanced effector cell stimulation by infected DCs was dependent on their secretion of interleukin 12. Immunization with infected DCs pulsed with tumor antigen protected against flank tumors in 78% of mice and induced a memory response. Antitumor immunity was dependent on both T cells and natural killer cells. Antigen-pulsed, mock-infected DCs were nonprotective. The findings that adenoviral vectors alone critically alter DC immune functions and antitumor properties have important implications for the design and interpretation of immunotherapy regimens using vector-based gene transfer to modulate immunity.

Biliary obstruction selectively expands and activates liver myeloid dendritic cells.

Obstructive jaundice is associated with immunologic derangements and hepatic inflammation and fibrosis. Because dendritic cells (DCs) play a major role in immune regulation, we hypothesized that the immunosuppression associated with jaundice may result from the functional impairment of liver DCs. We found that bile duct ligation (BDL) in mice expanded the myeloid subtype of liver DCs from 20 to 80% of total DCs and increased their absolute number by >15-fold. Liver myeloid DCs following BDL, but not sham laparotomy, had increased Ag uptake in vivo, high IL-6 secretion in response to LPS, and enhanced ability to activate T cells. The effects of BDL were specific to liver DCs, as spleen DCs were not affected. Expansion of liver myeloid DCs depended on Gr-1(+) cells, and we implicated monocyte chemotactic protein-1 as a potential mediator. Thus, obstructive jaundice selectively expands liver myeloid DCs that are highly functional and unlikely to be involved with impaired host immune responses.

Natural killer dendritic cells have both antigen presenting and lytic function and in response to CpG produce IFN-gamma via autocrine IL-12.

We have isolated rare cells bearing the NK cell surface marker NK1.1, as well as the dendritic cell (DC) marker CD11c, from the spleen, liver, lymph nodes, and thymus of normal mice. These cells possess both NK cell and DC function because they can lyse tumor cells and subsequently present Ags to naive Ag-specific T cells. Interestingly, in response to IL-4 plus either IL-2 or CpG, NKDC produce more IFN-gamma than do DC, or even NK cells. We determined that CpG, but not IL-2, induces NKDC to secrete IFN-gamma via the autocrine effects of IL-12. In vivo, CpG dramatically increases the number of NKDC. Furthermore, NKDC induce greater Ag-specific T cell activation than do DC after adoptive transfer. Their unique ability to lyse tumor cells, present Ags, and secrete inflammatory cytokines suggests that NKDC may play a crucial role in linking innate and adaptive immunity.

Conventional liver CD4 T cells are functionally distinct and suppressed by environmental factors.

The contribution of intrahepatic conventional T cells to the unique immunologic properties of the liver has not been clearly defined. We isolated bulk and CD4 T cells from mouse liver and compared their functions with each other and with their splenic counterparts. Unlike bulk spleen T cells, bulk liver T cells reacted minimally to allogeneic or antigen-loaded syngeneic dendritic cells. However, after exclusion of natural killer T cells (NKTs) and gammadelta T cells by FACS, liver and spleen CD4 T cells actually proliferated to a similar extent upon allogeneic or antigen-specific stimulation. Liver CD4 T cells were more sensitive to interleukin 2 (IL-2) than were spleen CD4 T cells, but had a similar proliferative potential based on their response to CD3 ligation. In addition, activated liver CD4 T cells produced higher levels of IL-4, IL-5, IL-10, and interferon gamma (IFN-gamma) than did splenic CD4 T cells. Therefore, liver CD4 T cells are intrinsically different from spleen CD4 T cells. In vitro, liver or spleen NKTs and gammadelta T cells suppressed liver and spleen CD4 T-cell proliferation in a dose-dependent fashion. In conclusion, unconventional T cells constrain liver CD4 T-cell function. Our findings have implications for pathological conditions of the liver that involve the response of conventional CD4 T lymphocytes.

GM-CSF expands dendritic cells and their progenitors in mouse liver.

Dendritic cells (DCs) are rare but ubiquitous antigen-presenting cells situated in lymphoid and nonlymphoid organs throughout the body. The study of DCs located in the liver has been restricted by their relative scarcity and the difficulty of their isolation. Because granulocyte-macrophage colony-stimulating factor (GM-CSF) is a critical growth factor for DCs in vitro, we postulated that it would expand hepatic DCs in vivo. We found that adenoviral-mediated GM-CSF overexpression in normal mice increased the number of liver DCs 400-fold to more than 100 million cells. GM-CSF-recruited DCs were CD11c(+)DEC205(-) and had high expression of major histocompatibility complex (MHC) class II, CD54, and CD80 but low CD40 and CD86 staining. Further maturation occurred after overnight culture. In addition to CD11c(+)DEC205(-) DCs, a population of CD11c(-)DEC205(low/-) cells resembling DC progenitors described previously in normal mice was expanded as serum GM-CSF levels increased. GM-CSF-recruited CD11c(+)DEC205(-) DCs and CD11c(-)DEC205(low/-) cells had different functional capabilities. CD11c(+)DEC205(-) DCs captured far more protein antigen in vivo, produced higher amounts of interleukin (IL)-6 and tumor necrosis factor (TNF)-alpha, and induced greater allogeneic and antigen-specific T-cell stimulation. A proportion of CD11c(-)DEC205(low/-) cells differentiated into CD11c(+) cells and gained T-cell stimulatory ability when cultured in the presence of GM-CSF. In conclusion, our findings show that GM-CSF can profoundly influence recruitment and development of DCs in murine liver.

Liver dendritic cells are less immunogenic than spleen dendritic cells because of differences in subtype composition.

The unique immunological properties of the liver may be due to the function of hepatic dendritic cells (DC). However, liver DC have not been well characterized because of the difficulty in isolating adequate numbers of cells for analysis. Using immunomagnetic bead and flow cytometric cell sorting, we compared freshly isolated murine liver and spleen CD11c+ DC. We found that liver DC are less mature, capture less Ag, and induce less T cell stimulation than spleen DC. Nevertheless, liver DC were able to generate high levels of IL-12 in response to CpG stimulation. We identified four distinct subtypes of liver DC based on the widely used DC subset markers CD8alpha and CD11b. Lymphoid (CD8alpha+CD11b-) and myeloid (CD8alpha-CD11b+) liver DC activated T cells to a similar degree as did their splenic DC counterparts but comprised only 20% of all liver DC. In contrast, the two more prevalent liver DC subsets were only weakly immunostimulatory. Plasmacytoid DC (B220+) accounted for 19% of liver DC, but only 5% of spleen DC. Our findings support the widely held notion that liver DC are generally weak activators of immunity, although they are capable of producing inflammatory cytokines, and certain subtypes potently activate T cells.

Optimization of dendritic cell maturation and gene transfer by recombinant adenovirus.

Dendritic cells (DC) have vast potential for immunotherapy. Transferring therapeutic genes to DC may enhance their inherent T cell-stimulatory capacity. Recombinant adenovirus is the most efficient vehicle for DC gene transfer and can alone mature DC. We sought to define the parameters of adenovirus infection of murine bone marrow-derived DC (BMDC) and the concomitant impact on BMDC maturation. The efficiency of adenoviral gene transfer to DC depended on the mouse strain, the organ source of DC, and the level of DC maturation. C57BL/6 BMDC consistently had higher transgene expression than BALB/c DC. While BMDC had considerable GFP expression after AdGFP infection, adenovirus was relatively ineffective in accomplishing transgene expression in freshly isolated hepatic or splenic DC. BMDC that were relatively immature because of a shorter duration of culture had higher transgene expression after infection. Nevertheless, pretreatment of DC with exogenous stimulants such as LPS or TNF-alpha resulted in higher transgene expression. Maturation of BMDC depended only on virus entry but not viral gene or transgene expression. Therefore, DC maturation was disproportionately high compared to the percentage of DC that actually expressed the adenoviral transgene. Maturation by adenovirus was only seen in BMDC, but not in liver or splenic DC, and was more pronounced in DC from later in culture (day 12 versus day 6). There was a dose-response relationship, up to a threshold dose, between adenovirus infection and both DC maturation and enhancement of DC activation of antigen-specific T cells. Our findings underscore the importance of optimizing gene transfer to DC in designing strategies for immunotherapy.

Increased and long-term generation of dendritic cells with reduced function from IL-6-deficient bone marrow.

The importance of IL-6 in dendritic cell (DC) development and function has not been well defined. To establish the role of IL-6, we studied bone marrow-derived DC (BMDC) and freshly isolated splenic DC from IL-6(-/-)-transgenic mice. We found that although IL-6(-/-) bone marrow had a similar composition to that of wild-type (WT) mice, it generated up to 10 times more DC when cultured in GM-CSF. The difference persisted even when IL-6(-/-) and WT bone marrow were cultured together, excluding the possibility that the effects were simply due to different cytokine microenvironments. In comparison to WT BMDC, IL-6(-/-) BMDC captured at least as much Ag, had an equivalent surface phenotype, and matured similarly in response to LPS or CpG. However, IL-6(-/-) BMDC induced less T cell allostimulation and Ag-specific T cell activation, but only the former was related to their inability to generate IL-6. Although WT bone marrow cultures died within 4 wk, IL-6(-/-) cultures continued to generate BMDC for >120 days, although the BMDC became immature and less functional. In vivo, we found that IL-6(-/-) mice had similar numbers and types of splenic DC as WT mice, both normally and after treatment with either Flt-3 ligand or GM-CSF. These findings demonstrate that IL-6 has profound effects on DC development in vitro, although the number and subtype composition of DC are unaffected by the absence of IL-6 in vivo. Furthermore, secretion of IL-6 is critical to certain DC functions.

Murine Flt3 ligand expands distinct dendritic cells with both tolerogenic and immunogenic properties.

Human Flt3 ligand can expand dendritic cells (DC) and enhance immunogenicity in mice. However, little is known about the effects of murine Flt3 ligand (mFlt3L) on mouse DC development and function. We constructed a vector to transiently overexpress mFlt3L in mice. After a single treatment, up to 44% of splenocytes became CD11c(+) and the total number of DC increased 100-fold. DC expansion effects lasted for >35 days. mFlt3L DC were both phenotypically and functionally distinct. They had increased expression of MHC and costimulatory molecules and expressed elevated levels of B220 and DEC205 but had minimal CD4 staining. mFlt3L DC also had a markedly altered cytokine profile, including lowered secretion of IL-6, IL-10, IFN-gamma, and TNF-alpha, but had a slightly increased capacity to stimulate T cells in vitro. However, in a variety of in vivo models, DC expanded by mFlt3L induced tolerogenic effects on T cells. Adoptive transfer of Ag-pulsed mFlt3L splenic DC to naive mice actually caused faster rates of tumor growth and induced minimal CTL compared with control DC. mFlt3L also failed to protect against tumors in which human Flt3 ligand was protective, but depletion of CD4(+) T cells restored tumor protection. Our findings 1) demonstrate that mFlt3L has distinct effects on DC development, 2) suggest an important role for mFlt3L in generating DC that have tolerogenic effects on T cells, and 3) may have application in immunotherapy in generating massive numbers of DC for an extended duration.

Natural killer cell depletion confounds the antitumor mechanism of endogenous IL-12 overexpression.

IL-12 gene transfer to hepatocytes using a recombinant adenovirus vector (AdIL-12) has been shown to protect against primary and metastatic liver tumors in mice. However, the mechanism of protection has been elusive and studies using depleting monoclonal antibodies or transgenic mice have purported it to be independent of T and NK cells. We postulated that depletion of NK cells may distort the experimental model and misrepresent the antitumor mechanism by altering the magnitude and duration of transgene expression. We show in mice treated with AdIL-12 that NK depletion increased serum IL-12 levels by more than 250-fold and prolonged transgene expression by nearly 2 weeks compared to nondepleted mice. To determine the contribution of NK cells to tumor protection after AdIL-12 treatment, we analyzed NK cells from treated animals. Isolated NK cells were markedly activated in terms of their lytic activity and IFN-gamma secretion. Adoptive transfer of NK cells from mice that had been treated with AdIL-12 to naive mice was sufficient to confer protection against colorectal hepatic metastases. This protection was mediated in part by NK-cell production of IFN-gamma. Our findings indicate that NK-cell depletion distorts the model of systemic AdIL-12 administration by markedly altering transgene expression, which then may potentiate other antitumor mechanisms, and that endogenous IL-12 overexpression activates NK cells, rendering them sufficient to protect against liver metastases. These data have critical implications for investigating the immunologic mechanisms of experimental models that utilize gene transfer.

Liver sinusoidal endothelial cells are insufficient to activate T cells.

Liver sinusoidal endothelial cells (LSEC) have been reported to express MHC class II, CD80, CD86, and CD11c and effectively stimulate naive T cells. Because dendritic cells (DC) are known to possess these characteristics, we sought to directly compare the phenotype and function of murine LSEC and DC. Nonparenchymal cells from C57BL/6 mice were obtained by collagenase digestion of the liver followed by density gradient centrifugation. From the enriched nonparenchymal cell fraction, LSEC (CD45(-)) were then isolated to 99% purity using immunomagnetic beads. Flow cytometric analysis of LSEC demonstrated high expression of CD31, von Willebrand factor, and FcgammaRs. However, unlike DC, LSEC had low or absent expression of MHC class II, CD86, and CD11c. LSEC demonstrated a high capacity for Ag uptake in vitro and in vivo. Although acetylated low-density lipoprotein uptake has been purported to be a specific function of LSEC, we found DC captured acetylated low-density lipoprotein to a similar extent in vivo. Consistent with their phenotype, LSEC were poor stimulators of allogeneic T cells. Furthermore, in the absence of exogenous costimulation, LSEC induced negligible proliferation of CD4(+) or CD8(+) TCR-transgenic T cells. Thus, contrary to previous reports, our data indicate that LSEC alone are insufficient to activate naive T cells.

Endogenous granulocyte-macrophage colony-stimulating factor overexpression in vivo results in the long-term recruitment of a distinct dendritic cell population with enhanced immunostimulatory function.

GM-CSF is critical for dendritic cell (DC) survival and differentiation in vitro. To study its effect on DC development and function in vivo, we used a gene transfer vector to transiently overexpress GM-CSF in mice. We found that up to 24% of splenocytes became CD11c+ and the number of DC increased up to 260-fold to 3 x 10(8) cells. DC numbers remained substantially elevated even 75 days after treatment. The DC population was either CD8alpha+CD4- or CD8alpha-CD4- but not CD8alpha+CD4+ or CD8alpha-CD4+. This differs substantially from subsets recruited in normal or Flt3 ligand-treated mice or using GM-CSF protein injections. GM-CSF-recruited DC secreted extremely high levels of TNF-alpha compared with minimal amounts in DC from normal or Flt3 ligand-treated mice. Recruited DC also produced elevated levels of IL-6 but almost no IFN-gamma. GM-CSF DC had robust immune function compared with controls. They had an increased rate of Ag capture and caused greater allogeneic and Ag-specific T cell stimulation. Furthermore, GM-CSF-recruited DC increased NK cell lytic activity after coculture. The enhanced T cell and NK cell immunostimulation by GM-CSF DC was in part dependent on their secretion of TNF-alpha. Our findings show that GM-CSF can have an important role in DC development and recruitment in vivo and has potential application to immunotherapy in recruiting massive numbers of DC with enhanced ability to activate effector cells.