Search

Increasing evidence supports the hypothesis that cancer stem cells (CSCs) are resistant to antiproliferative therapies, able to repopulate tumor bulk, and seed metastasis. NK cells are able to target stem cells as shown by their ability to reject allogeneic hematopoietic stem cells but not solid tissue grafts. Using multiple preclinical models, including NK coculture (autologous and allogeneic) with multiple human cancer cell lines and dissociated primary cancer specimens and NK transfer in NSG mice harboring orthotopic pancreatic cancer xenografts, we assessed CSC viability, CSC frequency, expression of death receptor ligands, and tumor burden. We demonstrate that activated NK cells are capable of preferentially killing CSCs identified by multiple CSC markers (CD24+/CD44+, CD133+, and aldehyde dehydrogenasebright) from a wide variety of human cancer cell lines in vitro and dissociated primary cancer specimens ex vivo. We observed comparable effector function of allogeneic and autologous NK cells. We also observed preferential upregulation of NK activation ligands MICA/B, Fas, and DR5 on CSCs. Blocking studies further implicated an NKG2D-dependent mechanism for NK killing of CSCs. Treatment of orthotopic human pancreatic cancer tumor-bearing NSG mice with activated NK cells led to significant reductions in both intratumoral CSCs and tumor burden. Taken together, these data from multiple preclinical models, including a strong reliance on primary human cancer specimens, provide compelling preclinical evidence that activated NK cells preferentially target cancer cells with a CSC phenotype, highlighting the translational potential of NK immunotherapy as part of a combined modality approach for refractory solid malignancies.

IL-27 exerts antitumor activity in murine orthotopic neuroblastoma, but only partial antitumor effect in disseminated disease. This study demonstrates that combined treatment with IL-2 and IL-27 induces potent antitumor activity in disseminated neuroblastoma metastasis. Complete durable tumor regression was achieved in 90% of mice bearing metastatic TBJ-IL-27 tumors treated with IL-2 compared with only 40% of mice bearing TBJ-IL-27 tumors alone and 0% of mice bearing TBJ-FLAG tumors with or without IL-2 treatment. Comparable antitumor effects were achieved by IL-27 protein produced upon hydrodynamic IL-27 plasmid DNA delivery when combined with IL-2. Although delivery of IL-27 alone, or in combination with IL-2, mediated pronounced regression of neuroblastoma metastases in the liver, combined delivery of IL-27 and IL-2 was far more effective than IL-27 alone against bone marrow metastases. Combined exposure to IL-27 produced by tumor and IL-2 synergistically enhances the generation of tumor-specific CTL reactivity. Potentiation of CTL reactivity by IL-27 occurs via mechanisms that appear to be engaged during both the initial sensitization and effector phase. Potent immunologic memory responses are generated in mice cured of their disseminated disease by combined delivery of IL-27 and IL-2, and depletion of CD8+ ablates the antitumor efficacy of this combination. Moreover, IL-27 delivery can inhibit the expansion of CD4+CD25+Foxp3+ regulatory and IL-17-expressing CD4+ cells that are otherwise observed among tumor-infiltrating lymphocytes from mice treated with IL-2. These studies demonstrate that IL-27 and IL-2 synergistically induce complete tumor regression and long-term survival in mice bearing widely metastatic neuroblastoma tumors.

The role of Fas in the regulation of solid tumor growth was investigated. Murine renal carcinoma (Renca) cells were constitutively resistant to Fas-mediated killing in vitro, but exhibited increased expression of Fas and sensitivity to Fas-mediated killing after exposure to IFN-γ and TNF. Transfected Renca cells overexpressing Fas were efficiently killed in vitro upon exposure to anti-Fas Ab (Jo2). When Fas-overexpressing Renca cells were injected into syngenic BALB/c mice, there was a consistent and significant delay in tumor progression, reduced metastasis, and prolonged survival that was not observed for Renca cells that overexpressed a truncated nonfunctional Fas receptor. The delay of in vivo tumor growth induced by Fas overexpression was not observed in IFN-γ−/− mice, indicating that IFN-γ is required for the delay of in vivo tumor growth. However, there was a significant increase of infiltrated T cells and in vivo apoptosis in Fas-overexpressing Renca tumors, and Fas-overexpressing Renca cells were also efficiently killed in vitro by T cells. In addition, a strong therapeutic effect was observed on Fas-overexpressing tumor cells by in vivo administration of anti-Fas Ab, confirming that overexpressed Fas provides a functional target in vivo for Fas-specific ligands. Therefore, our findings demonstrate that Fas overexpression on solid tumor cells can delay tumor growth and provides a rationale for therapeutic manipulation of Fas expression as a means of inducing tumor regression in vivo.

Mice bearing the experimental murine renal cancer Renca can be successfully treated with some forms of immunotherapy. In the present study, we have investigated the molecular pathways used by NK and T cells to lyse Renca cells. Renca cells normally express low levels of Fas that can be substantially enhanced by either IFN-γ or TNF-α, and the combination of IFN-γ + TNF-α synergistically enhances cell-surface Fas expression. In addition, cells pretreated with IFN-γ and TNF-α are sensitive to lysis mediated by Fas ligand (FasL)-expressing hybridomas (dllS), cross-linking of anti-Fas Abs or soluble Fas (FasL). Lysis via Fas occurs by apoptosis, since Renca shows all the typical characteristics of apoptosis. No changes in levels of bcl-2 were observed after cytokine treatments. We also examined cell-mediated cytotoxic effects using activated NK cells and T cells from gld FasL-deficient mice, and perforin-deficient mice, as well as wild-type C57BL/6 and BALB/c mice. Interestingly, the granule-mediated pathway predominated in killing of Renca by activated NK cells, while the Fas/FasL pathway contributed significantly to cell-mediated killing of Renca by activated T cells. These results suggest that killing of Renca tumor cells by immune effector cells can occur by both granule and Fas-mediated cytotoxicity. However, for the Fas-mediated pathway to function, cell surface levels of Fas need to be increased beyond a critical threshold level by proinflammatory cytokines such as IFN-γ and TNF-α.

Isolated murine splenic NK cells and the cultured murine endothelioma cell line, eEND2, were used to study the effects of cytokines on NK cell/endothelial cell adhesion. Treatment of eEND2 cells with TNF-alpha induced a marked increase (four- to sevenfold) in adherence of NK cells, as compared with control cultures of endothelioma cells or eEND2 cells treated with IL-1 alpha or IL-6. TNF-alpha induction of NK cell adherence to eEND2 was dose dependent with rapid kinetics, reaching a maximum at concentrations between 10 and 1000 U/ml after a 2-h incubation. TNF-alpha treatment of L929 fibroblasts or CL-2 hepatoma cells did not result in increased NK cell adhesion. The concentration range of TNF-alpha that was found to maximally augment NK cell adhesion to eEND2 also induced NK cell chemokinetic activity. The relevance of these in vitro results was subsequently analyzed in vivo. Initial studies confirmed that a single dose of polyinosinic-polycytidylic acid and poly-L-lysine stabilized in carboxymethyl cellulose (poly-ICLC), augmented hepatic NK activity and resulted in a 2.2-fold increase in the number of liver-associated NK cells. Concomitant treatment of mice with a TNF-alpha neutralizing antisera eliminated both the hepatic influx of NK cells and the increase in poly-ICLC-induced liver NK activity. These results suggest that TNF-alpha is a principal cytokine involved in the in vivo recruitment and localization of parenchymal NK cells after treatment with a biological response modifier, and that this regulation seems to occur via alterations in NK cell/endothelial cell interactions.

Human PBL (huPBL) were activated with anti-CD3 mAb in vitro and then were transferred into mice with severe combined immune deficiency (SCID) to determine the effect of activation on engraftment and to determine if the engrafted human cells could provide antitumor effects in mice. Some mice were also treated with human rIL-2 after huPBL transfer. Mice were analyzed 6 to 8 wk after cell transfer and the number of human cells in the peripheral lymphoid organs was determined. Mice receiving anti-CD3-activated huPBL demonstrated a significant increase in the incidence of human T cell engraftment in the periphery as assessed by flow cytometry. Human cells were also detected in the murine thymus after anti-CD3 stimulation indicating that human T cells can migrate to the murine thymus provided that they are activated before the transfer. The transfer of anti-CD3-activated huPBL also resulted in an xenogeneic graft-vs-host reaction manifested primarily by proliferation of murine splenic hemopoietic cells. When SCID mice received the human colon carcinoma HT29, the concurrent transfer of anti-CD3-activated human cells resulted in a significant increase in survival. However, the human cells displayed low cytolytic activity toward human tumor targets when they were recovered from the lymphoid organs of the SCID recipients. Supernatants from the anti-CD3-activated cells were able to inhibit the growth of HT29 in vitro, partly because of the presence of IFN-gamma, suggesting that the human T cells are producing cytokines in vivo that have antitumor effects. Thus, the use of anti-CD3-activated huPBL in SCID mice may be of value for optimizing human cell engraftment in the human/mouse lymphoid chimeras and may be used to evaluate potential anti-neoplastic therapies that employ human effector cells.

The resistance of tumors to cell death poses a major clinical problem. Bortezomib (Velcade/PS-341), a specific and reversible proteasome inhibitor, sensitized renal (Renca) and mammary (4T1) mouse carcinomas to apoptosis induced by the TNF family ligand TRAIL or its receptor (DR5) agonist antibody. The molecular basis of the tumoritoxicity of the combination of bortezomib and TRAIL seemingly involved amplification of the extrinsic cell death pathway as increased expression of DR5, activation of caspase-8 and -3, and decreased apoptosis inhibitory proteins cFLIP, cIAP-1 and XIAP were observed in bortezomib-treated tumors. In an established tumor setting, anti-DR5 administration to mice pre-treated with bortezomib not only reduced lung metastases of Renca and 4T1 tumors but also significantly increased long-term survival in mice bearing Renca as compared to either agent alone. Encouragingly, no evidence of any overt tissue damage was noted in the treated mice. This provides the first experimental evidence that proteasome inhibition can be combined with DR5 agonist antibody for successful cancer therapy. Moreover, bortezomib administration 24 hours prior to adoptive transfer of T cells did not block their anti-tumor effector function. This suggests that tumor sensitization with bortezomib may also be combined with tumor-specific T cell immunotherapy of cancer. Funded in part by NCI contract N01-CO-12400.

The i.p. injection of mice with highly purified recombinant human rIL-1 alpha or beta resulted in the rapid influx of a large number of polymorphonuclear neutrophils (PMN) into the peritoneal cavity. Significant increases in the number of PMN were induced by doses of IL-1 which ranged from 0.005 to 5 ng/injection. Interestingly the dose response for PMN influx was bell-shaped because 50 ng of IL-1 did not result in a significant increase in peritoneal PMN. IL-1 induced PMN infiltration was detectable by 1 h with peak levels of PMN obtained by about 2 h, followed by a subsequent decline by 24 h. Other cytokines, IL-2, IFN-gamma, IFN alpha beta, granulocyte-CSF, granulocyte-macrophage-CSF, IL-3, TNF-alpha, and TNF-beta were compared to IL-1 for their ability to induce a PMN influx into the peritoneum. Only TNF-alpha or TNF-beta (lymphotoxin) were able to induce a significant influx of PMN within 2 h. However, based on total protein administered, about 100 times more TNF than IL-1 was required to produce a comparable PMN infiltration. Intraperitoneal injection of inhibitors of the cyclooxygenase or lipoxygenase pathways did not inhibit the IL-1-induced influx of PMN. Also, neither IL-1 nor TNF triggered an increase in PG or leukotriene release from peritoneal cells in vitro. Furthermore, direct peritoneal injection of leukotriene B4, a potent PMN chemoattractant in vitro, did not induce any significant increase in PMN in the peritoneal cavity indicating that chemotactic activity alone is insufficient for inducing peritoneal infiltration. These results suggest that the local production of very low levels of IL-1 in vivo would be sufficient to initiate a sequence of events that results in a rapid accumulation of PMN. Because IL-1 was not chemotactic for PMN in vitro, our data suggest that IL-1 induces production of factors that are chemotactic for PMN. Alternatively, IL-1 may act on other stages of the complex sequence of events that regulates the emigration of PMN into tissue sites in vivo. The synergy apparent in PMN influx when suboptimal concentrations of IL-1 and TNF were injected suggests that the local production of very low concentrations of these cytokines in situ could play a critical role in the emigration of PMN during infection.

Lipid X, a monosaccharide biosynthetic precursor of lipid A, has been chemically synthesized and was shown to induce bone marrow-derived macrophages to release tumor necrosis factor (TNF) in vitro. However, relatively high amounts of lipid X were necessary for induction, and the levels of TNF were much less than those induced by small amounts of lipid A itself or LPS. Lipid X prepared by extraction of Escherichia coli mutants induced higher levels of TNF than the chemically synthesized material, but this is probably partially due to amounts of impurities in the extracted material. Pretreatment of macrophages with IFN-gamma resulted in the release of higher amounts of TNF on subsequent induction with either LPS or lipid X. In contrast, pretreatment of macrophages with LPS induced hyporesponsiveness for TNF production on subsequent rechallenge with LPS. Lipid X, on the other hand, was incapable of making macrophages hyporesponsive for TNF production.

Supernatants from the coculture of peripheral blood lymphocytes and the NK-susceptible cell line K562 were highly growth inhibitory for a variety of tumor cell lines. No correlation was observed between the susceptibility of the target cell lines to growth inhibition and to lysis by NK cells. Rather, the spectrum of cytostatic activity and the characteristics of the soluble factor were similar to those of leukoregulin, a recently described lymphokine. The supernatants of tumor-lymphocyte cultures contained only low levels of IFN-alpha and IFN-gamma, and antibodies to interferons did not affect the observed growth inhibition. The pattern of target cell susceptibility to growth inhibition by this factor was also quite distinct from that seen with purified recombinant LT or TNF. Furthermore, monoclonal antibodies to these cytokines also had no effect on the cytostasis, arguing against a requirement for, or synergistic interaction with, low levels of these cytokines. Some of the targets susceptible to the factor were only growth inhibited but not lysed, thereby distinguishing it from NKCF. Furthermore, the cytostasis was not inhibited by mannose-6-PO4 or rabbit antibodies to granule cytolysin, both of which have been reported to block NKCF. Therefore, the results show that a cytostatic factor is released in tumor-lymphocyte incubation that is quite distinct from interferons, LT, and TNF but has characteristics that resemble those of leukoregulin.

Natural killer cell activity was consistently increased after overnight incubation with recombinant IL 2. Recombinant IFN-gamma, on the other hand, increased NK activity only in three out of 25 preparations of donor lymphocytes. No synergy was observed when suboptimal amounts of recombinant (r)IL 2 and rIFN-gamma were added to donor lymphocytes, with any increase in activity attributable to additive effects of the two lymphokines. Three antibodies to IFN-gamma could not block the rIL 2 induction of NK activity, further suggesting that IFN-gamma was not involved in the enhancement of NK activity by IL 2. Two other anti-IFN-gamma antibody preparations showed significant inhibition of rIL 2-induced augmentation of NK activity, but the inhibition was found to be attributable to antibody-unrelated factors in the antiserum or ascites fluid. Our results suggest that IFN-gamma produced by rIL 2 treatment of human PBL does not play an essential role in increasing NK activity in most donors and that IL 2-induced augmentation of NK activity is due to the direct action of IL 2 on LGL.