Strategies for manipulating apoptosis for cancer therapy with tumor necrosis factor and lymphotoxin

Tumor necrosis factor (TNF) and lymphotoxin (LT), initially described as tumoricidal proteins, may be useful as adjuncts in cancer therapy. Treatment with TNF or LT was found to protect cells and animals against damage mediated by radiation or cytotoxic anticancer drugs. By contrast, tumor cells treated with TNF or LT were sensitized to these insults. We present a model in which TNF or LT induces both the synthesis of “protective” proteins such as manganous superoxide dismutase (MnSOD) and the activation of “killing” proteins, such as proteases, depending on the level of the inducing signal. Although the p55-TNF/LT receptor is structurally related to the Fas receptor, they can each signal apoptosis by distinct pathways. Furthermore, activation of both receptors acts synergistically in stimulating apoptosis. © 1996 Wiley-Liss, Inc.     

Cathepsin-G and leukocyte elastase inactivate human tumor necrosis factor and lymphotoxin

The addition of either cathepsin-G or leukocyte elastase to endotoxin-stimulated human peripheral blood monocytes decreased the immunoreactive tumor necrosis factor (TNF) detected in culture supernatants in a concentration-dependent manner. Both enzymes also induced a loss of supernatant cytolytic activity as determined on the WEHI-164 target cell line. Incubation of recombinant human TNF and lymphotoxin (LT) with either cathepsin-G or leukocyte elastase resulted in a loss of cytokine bioactivity. Examination of enzyme-treated recombinant cytokines by gel electrophoresis revealed that cathepsin-G cleaved LT into a 12.6-kDa fragment and leukocyte elastase fragmented LT into a 14.1-kDa product. On Western blots cathepsin-G and leukocyte elastase degraded TNF into 11- and 7.6-kDa fragments, respectively. Incubating leukocyte elastase with plasma elastase inhibitor alpha-1-antitrypsin prevented the loss of recombinant TNF bioactivity and blocked the degradation of this cytokine. This study suggests that two of the most abundant neutrophil proteases, cathepsin-G and leukocyte elastase, may be important regulators of TNF and LT bioactivity.     

Activation of cultured human endothelial cells by recombinant lymphotoxin: comparison with tumor necrosis factor and interleukin 1 species

Recombinant human lymphotoxin (LT) was compared with recombinant human tumor necrosis factor (TNF) for direct actions on cultured human endothelial cells (HEC). At equivalent half-maximal concentrations (based on L929 cytotoxicity units) LT and TNF each caused rapid and transient induction (peak 4 to 6 hr) of an antigen associated with leukocyte adhesion (detected by monoclonal antibody H4/18), a rapid but sustained increased expression (plateau 24 hr) of a lymphocyte adhesion structure (ICAM-1), a gradual (plateau 4 to 6 days) increase in expression of HLA-A,B antigens, and gradual (4 to 6 days) conversion of HEC culture morphology from epithelioid to fibroblastoid, an effect enhanced by immune interferon (IFN-gamma). Induction of H4/18 binding by maximal concentrations of LT or TNF could not be augmented by addition of the other cytokine, and 24 hr pretreatment with LT or TNF produced hyporesponsiveness to both mediators for reinduction. H4/18 binding can be transiently induced by tumor-promoting phorbol esters. Pretreatment with either LT or TNF also fully inhibited induction of H4/18 binding by phorbol ester, whereas phorbol ester pretreatment only variably and partially inhibited reinduction by LT or TNF. These actions of LT on endothelium shared with TNF may serve in vivo to promote lymphocyte and inflammatory leukocyte adhesion and transendothelial migration. Recombinant human interleukin 1 species (IL 1 alpha and IL 1 beta) shared many of the actions of LT and TNF and were indistinguishable from each other. However, IL 1 species could be distinguished from LT/TNF by their relative inability to enhance HLA-A,B expression, by their ability to augment H4/18 binding caused by maximally effective concentrations of LT or TNF, and by their inability to inhibit reinduction of H4/18 binding by LT or TNF. In contrast to the actions of LT or TNF, pretreatment with IL 1 alpha or IL 1 beta only partially inhibited induction of H4/18 binding by phorbol ester, and phorbol ester pretreatment consistently, albeit partially, inhibited induction by IL 1 species. These studies suggest that activated T cells through the secretion of LT can in turn activate the local endothelial lining so as to promote homing and extravasation of inflammatory cells. Furthermore, these LT actions can be augmented or complemented by other locally produced mediators such as IFN-gamma or IL 1.     

Studies on the differing effects of tumor necrosis factor and lymphotoxin on the growth of several human tumor lines

The relative ability of TNF and lymphotoxin (LT) to inhibit the growth of five human tumor cell lines was examined both in the presence and absence of IFN-gamma. Two adenocarcinoma lines, HT-29 and SK-CO-1, were 20- and 320-fold more sensitive to the inhibitory effects of TNF and LT in 3- to 4-day proliferation assays. In contrast, the breast carcinoma line BT-20 showed only a one- to twofold difference. The MCF-7 and ME-180 cell lines exhibited intermediate behavior. These results parallel the reported disparate potencies of TNF and LT in their effects on endothelial cells, hematopoietic development and their abilities to sustain a mixed lymphocyte response. Radiolabeled TNF binding studies showed two classes of receptors (Kd 0.04 to 0.15 nM and 0.2 to 1.0 nM) on the highly sensitive SK-CO-1 line. HT-29 cells also appeared to possess some high affinity-binding sites, whereas the BT-20 line completely lacked the high affinity form. Thus the presence of high affinity-binding sites correlated with increased sensitivity to the antiproliferative effects of TNF. Cold TNF competed with the binding of radiolabeled human TNF three- to fivefold better than LT for binding to all three lines. These relatively small differences between the TNF and LT receptor-binding characteristics are insufficient to explain the dramatic disparity in their antiproliferative properties. Likewise, the absolute concentrations of the unlabeled cytokines necessary to block the binding of 125I-TNF were 10- to 150-fold higher than the levels necessary to elicit the biologic response. Thus, the receptor binding data conflict with the growth inhibitory effects. This discrepancy is discussed in terms of either separate receptors for TNF and LT or more complex phenomena such as receptor cooperativity possibly resulting from multivalent interactions with the trimeric form of TNF.     

Expression of surface lymphotoxin and tumor necrosis factor on activated T, B, and natural killer cells.

The expression of membrane-associated forms of lymphotoxin (LT) and TNF were examined on cell lines of T, B, and myeloid origin, IL-2 dependent T cell clones, and peripheral blood lymphocytes. Inducible and constitutive patterns of surface LT expression were found on T cells as exemplified by the II-23.D7, a CD4+T cell hybridoma, and HUT-78, a T cell lymphoma. Phorbol ester induced surface LT expression on Ramos, an EBV transformed B cell line, but at a slower rate of appearance when compared to the II-23.D7. Secretion of LT was rapidly inducible by phorbol ester in II-23.D7 and also in HUT-78 but with slower kinetics; surface LT expression continued in both lines after secretion had ceased. Low levels of membrane TNF were transiently induced on II-23.D7 and HUT-78, but none was observed on Ramos. Peripheral blood monocytes and some myeloid tumor lines did not express surface LT. Several T cell clones expressed surface LT after Ag-specific stimulation, and expression persisted several days. Stimulation through the TCR or by IL-2 rapidly induced surface LT on resting peripheral T cells and CD56+ NK cells; pokeweed mitogen activation induced expression on CD20+ B cells. Consistent with previous results, immunoprecipitation with anti-LT mAb showed that LT was complexed with a distinct 33 kDa glycoprotein (p33) on cells that expressed surface LT, whereas secreted LT was not associated with p33. Surface and secreted modes of LT expression by activated T, B, and NK cells suggests that LT can be utilized as either a localized or diffusible mediator in immune responses.     

Tumor necrosis factor and lymphotoxin secretion by human natural killer cells leads to antiviral cytotoxicity

NK cells mediate their cytotoxicity against tumor cells through abroad array of cytotoxic and cytostatic proteins. We investigated whether specific proteins could also be identified that contributed to NK cell-mediated antiviral immunity. Human CD16+/CD3- NK cells were obtained by using FACS and subsequently cloned by using limiting dilution. These NK cell lines, which were cytotoxic against NK-sensitive tumor targets and virally infected cells, also generated supernatants that selectively killed vesicular stomatitis virus-infected cells while sparing noninfected cells. This soluble antiviral activity was completely neutralized by antibodies specific for TNF and lymphotoxin. Purified human rTNF also duplicated this specific cytotoxicity against vesicular stomatitis virus-infected cells, as well as against CMV-, Theiler's murine encephalomyelitis virus-, and HSV-infected cells. The degree of cytotoxicity varied for the different viruses and depended on the cell type infected. These results suggest that NK cells can mediate selective and direct cytotoxicity against virally infected cells by the secretion of TNF and lymphotoxin.     

Liposome-associated tumor necrosis factor retains bioactivity in the presence of neutralizing anti-tumor necrosis factor antibodies

Cell-associated TNF-alpha, either bound to its receptor on monocyte membranes or expressed as an integral membrane protein, can exert potent tumor cytolytic activity. We assessed the interaction of TNF with the lipid bilayer membrane system, liposomes, and the effects of membrane association on TNF bioactivity. High levels of TNF can be encapsulated within liposomes. At neutral pH, TNF binds to the surface of preformed liposomes (liposome-associated TNF), but does not partition into the lipid bilayer. TNF appears to bind to negatively charged phospholipid head groups of the outer membrane leaflet. Free TNF, liposome-associated TNF, and liposome-encapsulated TNF display comparable abilities to activate human peripheral blood monocytes and to lyse tumor cells. However, liposome-encapsulated TNF, as well as TNF bound to the outer surface of preformed liposomes, retains bioactivity in the presence of anti-TNF antibodies that neutralize free TNF. The interaction of liposomal TNF with cell surface TNF receptors thus appears to be preserved in the presence of neutralizing antibodies.     

Microtubule plus-end-tracking proteins target gap junctions directly from the cell interior to adherens junctions

Gap junctions are intercellular channels that connect the cytoplasms of adjacent cells. For gap junctions to properly control organ formation and electrical synchronization in the heart and the brain, connexin-based hemichannels must be correctly targeted to cell-cell borders. While it is generally accepted that gap junctions form via lateral diffusion of hemichannels following microtubule-mediated delivery to the plasma membrane, we provide evidence for direct targeting of hemichannels to cell-cell junctions through a pathway that is dependent on microtubules; through the adherens-junction proteins N-cadherin and beta-catenin; through the microtubule plus-end-tracking protein (+TIP) EB1; and through its interacting protein p150(Glued). Based on live cell microscopy that includes fluorescence recovery after photobleaching (FRAP), total internal reflection fluorescence (TIRF), deconvolution, and siRNA knockdown, we propose that preferential tethering of microtubule plus ends at the adherens junction promotes delivery of connexin hemichannels directly to the cell-cell border. These findings support an unanticipated mechanism for protein delivery to points of cell-cell contact.     

Tumor necrosis factor stimulates ornithine decarboxylase activity in human fibroblasts and tumor target cells

The activity of the polyamine biosynthetic enzyme, ornithine decarboxylase (ODC), has been shown to be rapidly modulated by a variety of growth regulatory molecules. In this report the effect of the growth modulatory peptide, tumor necrosis factor, on ODC activity was examined on two cell lines which express equivalent TNF binding properties, but differ in their growth response when exposed to this factor. TNF treatment of WI-38 fibroblasts stimulated both their growth and induced ODC activity 5-10-fold when measured 6-24 h after TNF incubation. TNF induced cytotoxicity in ME-180 cervical carcinoma cells and, interestingly, stimulated both ODC activity (3-6-fold) and putrescine accumulation when measured prior to the onset of cytotoxicity. Induction of ODC was TNF concentration-dependent and paralleled the concentration-dependency for cytotoxicity. Based upon studies with cycloheximide, de novo protein biosynthesis was required for TNF-mediated ODC induction in ME-180 cells. The effects of other growth inhibitory peptides and growth factors were analyzed for their combined effect on ODC activity in TNF-treated or untreated ME-180 cells. Interferon gamma treatment had no significant effect on basal ODC activity but inhibited TNF-mediated ODC induction by approximately 50%. EGF treatment resulted in a potent stimulation of ODC activity which was not affected by TNF pre-treatment or coadministration on ME-180 cells. These results suggest that TNF has properties which are similar to those of a growth factor and distinct from those of other growth inhibitory peptides. The early growth factor-like actions of TNF occur on both normal fibroblasts and some tumor cells and evidence suggests that these effects are antagonistic to the antiproliferative effects of TNF.     

Human lymphotoxin and tumor necrosis factor genes: structure, homology and chromosomal localization.

Human Tumor Necrosis Factor and Lymphotoxin are cytotoxic proteins which have similar biological activities and share 30 percent amino acid homology. The single copy genes which encode these proteins share several structural features: each gene is approximately three kilobase pairs in length and is interrupted by three introns. In addition, these genes are closely linked and have been mapped to human chromosome 6. However, only the last exons of both genes, which code for more than 80 percent of each secreted protein, are significantly homologous (56 percent).     

Synergistic interactions between interleukin 1, tumor necrosis factor, and lymphotoxin in bone resorption

Cytokines with bone-resorbing activity include IL 1 beta (pI 7), IL 1 alpha (pI 5), tumor necrosis factor (TNF), and lymphotoxin (LT). Possible interaction between IL 1 beta, the major mediator with osteoclast-activating factor (OAF) activity, and other cytokines was studied. By itself, IL 1 beta was 13-fold more potent than IL 1 alpha and 1000-fold more potent than either TNF or LT in stimulating bone resorption. Suboptimal concentrations of IL 1 beta or IL 1 alpha in combination with suboptimal concentrations of TNF or LT resulted in synergistic bone-resorptive responses (1.5 to 10 times the expected responses if their effects were additive). Synergy between either form of IL 1 and TNF or LT resulted in a twofold increase in activity of IL 1, and a 100-fold increase in activity of TNF or LT. However, even with optimal synergy, IL 1 beta remained 20-fold more potent in inducing bone resorption than TNF or LT. Because IL 1 beta is considerably more potent than TNF and LT in stimulating bone resorption either alone or under synergistic conditions, it is unlikely that TNF and LT are responsible for more than a minor proportion of the total bone-resorbing activity formerly referred to as OAF.     

Analysis of a cytostatic lymphokine produced by incubation of lymphocytes with tumor cells: relationship to leukoregulin and distinction from recombinant lymphotoxin, recombinant tumor necrosis factor, and natural killer cytotoxic factor

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.     

Two conserved tryptophan residues of tumor necrosis factor and lymphotoxin are not involved in the biological activity

Each of the two highly conserved tryptophan residues in hTNF (positions 28 and 114) was converted into phenylalanine by site-directed mutagenesis and the mutant proteins were partially purified. A cytotoxicity assay on mouse L929 cells showed only a slight reduction in biological activity, strongly suggesting that neither of the two amino acids is involved in the active site.     

Resistance of tumor cells to cytolytic T lymphocytes involves Rho-GTPases and focal adhesion kinase activation

Tumor cells evade adaptive immunity by a variety of mechanisms, including selection of variants that are resistant to specific cytotoxic T lymphocyte (CTL) pressure. Recently, we have reported that the reorganization of the actin cytoskeleton can be used by tumor cells as a strategy to promote their resistance to CTL-mediated lysis. In this study, we further examined the functional features of a CTL-resistant tumor variant and investigated the relationship between cytoskeleton alteration, the acquisition of tumor resistance to CTL-induced cell death, Rho-GTPases, and focal adhesion kinase (FAK) pathways. Our data indicate that although the resistant cells do not display an increased migratory potential, an alteration of adhesion to the extracellular matrix was observed. When Rho-GTPases were activated in cells by the bacterial CNF1 (cytotoxic necrotizing factor 1), striking changes in the cell morphology, including actin cytoskeleton, focal adhesions, and membrane extensions, were observed. More importantly, such activation also resulted in a significant attenuation of resistance to CTL-induced cell death. Furthermore, we demonstrate that FAK signaling pathways were constitutively defective in the resistant cells. Silencing of FAK in the sensitive target cells resulted in the inhibition of immune synapse formation with specific CTLs and their subsequent lysis. Expression of the FAK mutant (Y397F) resulted in an inhibition of IGR-Heu cell adhesion and of their susceptibility to specific lysis. These results suggest that FAK activation plays a role in the control of tumor cell susceptibility to CTL-mediated lysis.     

Targeting of tumor necrosis factor to tumor cells: secretion by myeloma cells of a genetically engineered antibody-tumor necrosis factor hybrid molecule

The construction, synthesis and secretion of a genetically engineered antibody-cytokine fusion molecule is described. To target tumor necrosis factor (TNF) to tumor cells, recombinant antibody techniques were used to produce a Fab-like antibody-TNF conjugate. At the gene level, the heavy chain gene of an antitransferrin receptor antibody was linked to a synthetic TNF gene encoding human TNF. Transfection of the heavy chain-TNF gene into a myeloma derived cell line which was producing the light chain of the same antibody, allowed the isolation of a cell line secreting a fusion protein of the expected molecular weight and composition. The culture supernatant of the cell line contained TNF cytotoxic activity towards murine L929 cells and human MCF-7 cells. Cytotoxicity towards the human cancer cells was inhibited by an excess of the original antitransferrin receptor antibody, indicating that the antibody-TNF molecules are targeted to the transferrin receptor rich tumor cells. Since the antibody genes used are chimeric (i.e. composed of mouse variable and human constant regions) and since DNA encoding human TNF was used, the hybrid protein is an example of a humanized immunotoxin-like molecule. These results illustrate the possibilities of antibody engineering technology to create and produce improved agents for cancer therapy. Furthermore, they demonstrate for the first time the ability of myeloma cells to secrete an antibody-cytokine chimeric molecule.