Interferon-alpha down-regulates insulin receptors in lymphoblastoid (Daudi) cells. Relationship to inhibition of cell proliferation

The Daudi line of human lymphoblastoid cells requires insulin and transferrin for growth in serum-free medium and is highly sensitive to the inhibitory effect of human leukocyte interferon (IFN-alpha) on cell proliferation. A variant subline of Daudi cells, which is resistant to the antiproliferative action of IFN-alpha, also has been grown in serum-free medium containing insulin and transferrin. The proliferation of IFN-sensitive and -resistant Daudi cells is dependent on the occupancy of insulin receptors, with optimal cell proliferation observed at high receptor occupancy (nearly 100%). No evidence was found for receptors for insulin-like growth factor I on Daudi cells. IFN treatment of IFN-sensitive cells decreased the capacity of the cells to bind 125I-insulin. The altered binding capacity was due to diminished specific, lower affinity insulin binding, as detected at high 125I-insulin concentrations. Higher affinity insulin binding was not altered by IFN. Insulin binding was also reduced in detergent-solubilized extracts from IFN-treated sensitive Daudi cells and the magnitude of the effect was comparable to that observed in intact cells. This indicates that the total number of insulin binding sites (surface + internal) is decreased in IFN-treated sensitive cells. Insulin binding to IFN-sensitive cells decreased linearly with time between 6 and 48 h from the addition of IFN. The effect on lower affinity insulin binding developed more rapidly than the inhibitory effect of IFN on cell proliferation. The insulin-binding capacity of Daudi cells resistant to the antiproliferative effect of IFN was unaffected by IFN, despite the fact that these cells contain as many cell surface IFN receptors as sensitive cells. These observations raise the possibility that lower affinity insulin binding is important in the growth-promoting actions of insulin.     

Therapy-induced antibodies against the antiviral and antiproliferative effects of interferons in patients with chronic hepatitis C virus infection

Sera from 86 patients with chronic hepatitis C virus (HCV) infection treated with recombinant interferons-alpha (rIFN-alpha) were screened for IFN-binding and antiviral effect-neutralizing antibodies. Out of the 61 patients treated with rIFN-alpha2b, 46% had binding and 28% had neutralizing antibodies. 44% of the 25 patients treated with rIFN-alpha2a developed binding antibodies and 24% had neutralizing antibodies. Contradictory data were observed concerning the appearance of anti-IFN antibodies and the outcome of IFN therapy. A significantly higher number of the patients with a sustained response to rIFN-alpha2b therapy formed antibodies than the number among the non-responder patients. At the same time, in the patients treated with rIFN-alpha2a, opposite data were found. The activity of the antibodies in some sera was studied against the antiproliferative effect of IFNs on Daudi cells by measuring the [3H]thymidine incorporation. The binding antibodies without neutralization of the antiviral effect of the IFNs inhibited the antiproliferative activity of the rIFNs, similarly to antibodies having both IFN-binding and antiviral effect-neutralizing capacities. At the same time, the antiproliferative effect of the natural IFN was less affected. It is suggested that the antiproliferative assay is more sensitive than the antiviral method for demonstration of the presence of antibodies exerting an inhibitory effect on the biological activities of IFN.     

Comparative antiproliferative and receptor binding activities of interferons alpha and beta on lymphoblastoid and melanoma cells

The relative antiproliferative and receptor binding characteristics of the hitherto little-characterized interferon alpha 4a on cells of lymphoid and epithelial origin are compared with two other type I interferons, alpha 2 a and beta. Using the lymphoblastoid cell line, Daudi, interferons alpha 4 a and alpha 2 b had similar antiproliferative activity, and were about 10-fold more active than IFN beta. By contrast, using the melanoma cell line Sk-Mel-28, IFN beta was the most active, whereas IFN alpha 2b and IFN alpha 4a were respectively 60-fold and greater than 1000-fold less active than on Daudi cells. Receptor binding did not correlate with antiproliferative sensitivities, but confirmed a shared receptor component for these three interferons. These results indicate that the antiproliferative activities of three type I IFNs differs markedly on different cell types and that this is unlikely to be due to receptor binding, but more likely a post receptor binding event.     

Effects of recombinant interferons on the clonogenic growth of leukemic cells and normal hemopoietic progenitors

We have examined the effects of recombinant immune and leukocyte interferons (rIFN-gamma and rIFN-alpha) on the clonogenic growth of leukemic cells and normal hemopoietic progenitors using in vitro colony assays. Both interferons suppressed the colony formation by granulocyte-macrophage progenitors (CFU-gm) and erythroid progenitors (CFU-e and BFU-e) in a dose-dependent manner. Six myeloid leukemic cell lines were less sensitive to rIFN-gamma than CFU-gm. The colony formation of some myeloid leukemic cell lines was suppressed more potently by rIFN-alpha than by CFU-gm. Four lymphoid leukemic cell lines of the T-cell type were very resistant to both recombinant interferons. Reduced sensitivity of leukemic cells to rIFN-gamma, a possible hemopoietic regulator, may explain partially the unregulated proliferation of leukemic cells in vivo.     

Kaposi's sarcoma-associated herpesvirus viral interferon regulatory factor confers resistance to the antiproliferative effect of interferon-alpha

BACKGROUND: Kaposi's sarcoma-associated herpesvirus (KSHV) encodes a 442 amino acid polypeptide-designated viral interferon regulatory factor (vIRF) that displays homology to members of the interferon regulatory factor (IRF) family that bind to consensus interferon sequences and transactivate cellular genes that can modulate growth inhibition. Studies were conducted to determine whether vIRF affects the growth suppression mediated by interferon-alpha (IFN-alpha) in a human B lymphocyte cell line. MATERIALS AND METHODS: The human B lymphocyte cell line Daudi, which is sensitive to the antiproliferative effects of IFN-alpha, was stably transfected to express vIRF, and the proliferative response of vIRF expressing cells to IFN-alpha was compared with controls. The effect of vIRF on IRF- 1 transactivation was analyzed by co-transfection of an IFN-alpha-responsive chloramphenicol acetyltransferase reporter and expression plasmids encoding IRF-1 and vIRF. Electrophoretic mobility shift assays were conducted to determine whether vIRF interferes with the DNA binding activity of IRF-1. RESULTS: Daudi human B lymphocyte cells expressing vIRF were resistant to the antiproliferative effects of IFN-alpha, whereas wild-type Daudi or Daudi cells transformed with vector DNA were growth inhibited by IFN-alpha. The activation of an interferon-responsive reporter by IFN-alpha or IRF-1 was repressed by expression of vIRF. IRF-1 DNA binding activity was unaffected by vIRF, and vIRF alone did not bind to the interferon consensus sequence. CONCLUSIONS: These studies revealed that vIRF functions to inhibit interferon-mediated growth control of a human B lymphocyte cell line by targeting IRF-1 transactivation of interferon-inducible genes. Since KSHV is a B lymphotropic herpesvirus associated with two forms of B lymphocyte neoplasms, these effects of vIRF likely contribute to B cell oncogenesis associated with KSHV infection.     

Transforming growth factor-beta 1 enhances the suppression of human hematopoiesis by tumor necrosis factor-alpha or recombinant interferon-alpha

The effects of transforming growth factor-beta 1 (TGF-beta 1) on human hematopoiesis were evaluated in combination with two other regulatory cytokines, namely, recombinant human tumor necrosis factor-alpha (TNF-alpha) and recombinant human interferon-alpha (rIFN-alpha). Combinations of TNF-alpha and TGF-beta 1 resulted in a synergistic suppression of colony formation by erythroid progenitor cells (BFU-E) and an additive suppression of granulocyte-macrophage (CFU-GM) and multipotential (CFU-GEMM) progenitor cells. In addition, TGF-beta 1 synergized with rIFN-alpha to suppress CFU-GM formation, while the combined suppressive effects of both cytokines on CFU-GEMM and BFU-E were additive. When TGF-beta 1 was tested with TNF-alpha or IFN-alpha on granulocyte/macrophage colony-stimulating factor (GM-CSF)-stimulated bone marrow cells in a 5-day proliferation assay, the antiproliferative effects of TGF-beta 1 and TNF-alpha were additive, while those with TGF-beta 1 and rIFN-alpha were synergistic. A similar pattern was seen in the suppression of the myeloblastic cell line KG-1 where TGF-beta 1 in combination with TNF-alpha resulted in an additive suppression while inhibition by TGF-beta 1 and IFN-alpha was synergistic. These results demonstrate for the first time the cooperative effects between TGF-beta and TNF-alpha and IFN-alpha in the suppression of hematopoietic cell growth, raising the possibility that TGF-beta might be used in concert with TNF-alpha or IFN-alpha in the treatment of various myeloproliferative disorders.     

Signal Transmission through MHC Class II Molecules in a Human B Lymphoid Progenitor Cell Line: Different Signaling Pathways Depending on the Maturational Stages of B Cells

The function of MHC class II HLA-DR molecules expressed on a human B lymphoid progenitor cell line FL8.2.4.4 (abbreviated as FL4.4) was examined. FL4.4 cells expressed HLA-DR molecules and stimulation of the DR molecules by anti-DR mAb or by superantigen TSST-1 induced strong augmentation of homocytic aggregation and protein tyrosine phosphorylation in FL4.4 cells. Induced homocytic aggregation in FL4.4 consists both of LFA-1/ICAM-1-dependent and -independent pathways as revealed by mAb blocking experiments. Metabolic inhibitors, NaN3 and cytochalasin B, blocked the induced homocytic aggregation of FL4.4. Early mature Daudi B cell lines also showed a similar type of homocytic aggregation by stimulation with anti-DR mAb. Daudi cells are more sensitive to protein kinase inhibitors herbimycin A and H7 than FL4.4 cells in their blocking of induced homocytic aggregation, while W7 showed stronger inhibitory effects on FL4.4 cells than on Daudi cells. Western blotting analysis revealed that the stimulation of DR molecules induced protein tyrosine phosphorylation of 100-kDa, 90-kDa, 60-kDa and 55-kDa proteins in FL4.4 cells, while, in Daudi cells 110-kDa, 100-kDa and 80-kDa proteins were phosphorylated. These results suggest that different signaling pathways through class II molecules are employed depending on the maturational stage of B-cell differentiation.     

Early plasma membrane depolarization by alpha interferon: biologic correlation with antiproliferative signal

Daudi lymphoma cells, of a line sensitive to growth inhibition by alpha interferon, showed dose-dependent plasma membrane depolarization within 10 min after exposure to natural or recombinant alpha interferons (10 to 1000 IU/ml). This biophysical change was detected flow cytometrically by measuring the intensity of fluorescent emission from cells stained with dye indicators of membrane potential. Subclones of Daudi lymphoma cells, resistant to growth inhibition by alpha interferon, showed no membrane depolarization. Parallel results were obtained in initial tests of an isologous pair of T cell and B cell lines which differ in sensitivity to growth inhibition. Thus, decreased membrane potential may herald an interferon signal for antiproliferative action.     

Differences in sialyl transferase activity and in concanavalin-A agglutination between T and B lymphoblastoid cell lines.

The concanavalin A agglutination patterns, sialyl transferase activity and sialic acid content were studied for cultured lymphoblastoid cell lines possessing either T or B surface markers. Concanavalin A caused marked agglutination of the two B cell lines studied, Raji and SB, while the two T cell lines, HSB-2 and CCRF-CEM, failed to agglutinate with this lectin. The surface sialyl transferase activity of the two B lines was significantly higher than on the two T lines studied. In contrast, the total cellular sialic acid content or the surface sialic acid that was released from the T and B cell lines by neuraminidase was not significantly different. This study indicates that T and B lymphoblastoid cells possess different levels of surface located sialyl transferase activity and display different agglutination patterns with Con A. Perhaps these assays can be of value in differentiating various classes of neoplastic lymphoid cells.     

Single high affinity binding of interferon α2 to receptors on human lymphoblastoid cells: Internalization and inactivation of receptors

The interaction of human recombinant interferon (rIFN) α2 with its receptor on lymphoblastoid cells was studied using competitive displacement binding. The data were analysed with the LIGAND program, which tests their fit to one-site or multiple binding site models. The binding at 4° and 37°C fits a one-site model, with a similar K_D for both IFN-sensitive and resistant cells. Binding at 37°C to Daudi cells at high density fits artifactually a two-site mode only when the receptor concentration is close to that of the K_D. The binding of IFN to its receptor, therefore, follows a simple bimolecular interaction. Furthermore, IFN-sensitive and resistant cells internalize IFN at similar rates. We have examined whether IFN receptors are also internalized and whether they subsequently recycle to the cell surface. By measuring cell surface and total receptors, we have observed that after 2 h treatment with IFN total receptors remain constant whereas cell surface receptors decrease. After prolonged treatment with IFN, however, there is a loss of total receptors. By inactivating cell surface receptors with proteinase K, we have shown that a fraction of cell surface receptors becomes resistant to inactivation and is apparently internalized. Moreover, experiments which measure IFN receptors either during incubation in the presence of IFN or after IFN has been removed from the medium, show that receptors do not recycle to the cell surface after internalization. The addition of monensin, a drug which has been shown to inhibit receptor recycling, has no effect on the loss of IFN receptors.     

Immunochemotherapy of malignant glioma: synergistic activity of CD95 ligand and chemotherapeutics

 Malignant glioma cells are susceptible to CD95(Fas/APO-1)-mediated apoptosis triggered by agonistic antibody. Here we examined the proapoptotic effects of the natural CD95 ligand, a cytotoxic cytokine homologous to tumor necrosis factor, on malignant glioma cell lines LN-229, LN-308 and T98G. We assessed whether glioma cell killing is synergistically enhanced by cotreatment with CD95 ligand and chemotherapeutic agents, including doxorubicin, carmustine, vincristine, etoposide, teniposide, 5-fluorouracil and cytarabine. Synergy was examined at low concentrations of cytotoxic drugs and CD95 ligand with a defined effect level (IC₁₅). Short-term-cytotoxicity assays showed prominent killing of the glioma cells by CD95 ligand but not by the drugs at relevant concentrations. CD95 ligand-induced apoptosis in the acute toxicity paradigm was augmented by doxorubicin and vincristine. Growth-inhibition assays revealed prominent synergy between CD95 ligand and all drugs examined. The best synergy was obtained with CD95 ligand and doxorubicin, vincristine or teniposide. The strong synergistic antiproliferative effects were observed at much lower concentrations of CD95 ligand and cytotoxic drugs than the moderate synergistic acute cytotoxic effects. All cell lines examined express the Bcl-2 protein. LN-229 has partial wild-type p53 activity. T98G has mutant p53. LN-308 has a deleted p53 gene and lacks p53 protein expression. Thus, synergistic effects of CD95 ligand and cytotoxic drugs were observed in cell lines exhibiting two features thought to play a role in the chemoresistance of human malignant glioma cells: loss of wild-type p53 activity and acquisition of bcl-2 expression. Ectopic expression of murine bcl-2 conferred partial protection from CD95 ligand and drugs when administered alone but did not interfere with the mechanisms underlying the synergistic effects of CD95 ligand and chemotherapeutic drugs. Received: 31 October 1996 / Accepted: 4 January 1997     

Spontaneous membrane transfer through homotypic synapses between lymphoma cells

Formation of an immunological synapse by T, B, or NK cells is associated with an intercellular transfer of some membrane fragments from their respective target cells. This capture is thought to require effector cell activation by surface recognition of stimulatory ligand(s). However, spontaneous synaptic transfers between homotypic lymphoid cells has never been described. In this study, we show that without adding Ag, resting healthy lymphoid cells and several tumor cell lines are inactive. Conversely, however, some leukemia cell lines including the Burkitt's lymphoma Daudi continuously uptake patches of autologous cell membranes. This intercellular transfer does not involve cytosol molecules or exosomes, but requires cell contact. In homotypic Daudi cell conjugates, this occurs through immunological synapses, involves constitutive protein kinase C and mitogen-activated protein/extracellular signal-regulated kinase kinase activity and strongly increases upon B cell receptor activation. Thus, spontaneous homosynaptic transfer may reflect the hitherto unsuspected autoreactivity of some leukemia cell lines.     

Role of the production of natural killer cell stimulatory factor (NKSF/IL-12) in the ability of B cell lines to stimulate T and NK cell proliferation.

We have previously used human Epstein Barr virus (EBV)-transformed B lymphoblastoid cell lines for the identification and purification of a novel cytokine, natural killer cell stimulatory factor (NKSF/IL-12), that has pleiotropic effects on human lymphocytes. B cell lines are also routinely employed as feeder cells for the culture of T and natural killer (NK) cells. In this report we describe the ability of two NKSF/IL-12 producing B cell lines (RPMI-8866 and Cess) and two nonproducing lines (Raji and Daudi) to stimulate the proliferation of T and NK cells in 8-day PBL cultures. We demonstrate, using an anti-NKSF/IL-12 neutralizing monoclonal antibody, that the endogenous production of NKSF/IL-12 in these cultures can significantly enhance the proliferation and cytotoxic activity of T and NK cells. We also report that the addition of exogenous rNKSF/IL-12 can greatly increase the number of T and NK cells obtained from the cultures following stimulation by the B cell lines. Aside from the possible practical applications, the enhanced proliferation of T and NK cells consistently observed in the presence of endogenously produced NKSF/IL-12 or exogenously added rNKSF/IL-12 in this system may further our understanding of the role of this cytokine during an in vivo immune response.     

Biological Properties of Melanoma and Endothelial Cells after Plasmid AMEP Gene Electrotransfer Depend on Integrin Quantity on Cells

The data on the biological responsiveness of melanoma and endothelial cells that are targeted by Antiangiogenic MEtargidin Peptide (AMEP) are limited; therefore, the antiproliferative, antimetastatic and antiangiogenic effects of AMEP were investigated in murine melanoma and human endothelial cells after plasmid AMEP gene electrotransfer into the cells in vitro. Plasmid AMEP, a plasmid coding for the disintegrin domain of metargidin targeting specific integrins, had cytotoxic and antiproliferative effects on murine melanoma and human endothelial cells. Among the metastatic properties of cells, migration, invasion and adhesion were investigated. Plasmid AMEP strongly affected the migration of murine melanoma and human endothelial cell lines and also affected the invasion of highly metastatic murine melanoma B16F10 and human endothelial cell lines. There was no effect on cell adhesion on Matrigel^TM or fibronectin in all cell lines. The antiangiogenic effect was shown with tube formation assay, where human microvascular endothelial cell line (HMEC-1) proved to be more sensitive to plasmid AMEP gene electrotransfer than the human umbilical vein endothelial cell line (HUVEC). The study indicates that antiproliferative and antimetastatic biological responses to gene electrotransfer of plasmid AMEP in murine melanoma cells were dependent on the integrin quantity on melanoma cells and not on the expression level of AMEP. The strong antiangiogenic effect expressed in human endothelial cell lines was only partly dependent on the quantity of integrins and seemed to be plasmid AMEP dose dependent.     

Binding of human tumor necrosis factor to high affinity receptors on HeLa and lymphoblastoid cells sensitive to growth inhibition

Purified human tumor necrosis factor (TNF) was iodinated to high specific activity with good retention of its biological activity, as determined by the cytotoxic titer on murine L929 cells. The binding of 125I-TNF to L929 and human HeLa S2 cells grown in monolayer was initially measured, but high levels of nonspecific binding were observed. Specific binding to high affinity receptors of HeLa S2 cells grown in suspension culture was demonstrated by competitive displacement experiments and analysis of the binding data with the LIGAND program. A KD of 2 X 10(-10) M and 6000 receptors/cell were calculated in this way. These observations provide the first direct evidence for a cellular receptor for TNF. The cell-bound 125I-TNF was internalized at 37 degrees C, presumably by receptor-mediated endocytosis, and subsequently degraded to acid-soluble products. Three lines of human lymphoblastoid cells were examined for sensitivity to the cytostatic effect of TNF and for the presence of high affinity receptors. Daudi and Raji cells were insensitive to TNF and showed very few specific binding sites when incubated with 125I-TNF. Jurkat cells were growth-inhibited by TNF and showed a significantly greater number of specific binding sites than the other lymphoblastoid cells. These findings suggest that the sensitivity of some cell lines to the biological effects of TNF may be correlated with the presence of a relatively high number of receptors for this factor.     

SENSITIVITY OF DAUDI CELLS TO TUMOR NECROSIS FACTOR ALPHA: EARLY CHANGES IN POLYPHOSPHOINOSITIDE METABOLISM

The effects of recombinant Tumor Necrosis Factor α (r-TNF α) on polyphosphoinositide metabolism were examined in a Burkitt Lymphoma cell line (Daudi cells). After 1h of in vitro treatment with r-TNF α, the incorporation of³²Pi into phosphatidylinositol 4,5-phosphate (PtdInsP₂), phosphatidylinositol 4-phosphate (PtdInsP) and phosphatidylinositol (PtdIns) was reduced compared with controls, confirming previous findings observed in other cell lines of a specific PtdIns breakdown following r-TNF α treatment. The novelty of this study is therefore the demonstration of early changes in polyphosphoinositide metabolism during the antiproliferative response elicited by this cytokine in Daudi cells.     

A monoclonal antibody to human B lymphoblastoid cells activates human and murine T lymphocytes

A B lymphoblastoid cell line can provide a comitogenic, accessory signal for mitogen-treated T cells. In a study evaluating the antigenic determinant of such cells that mediate this effect, a monoclonal antibody (I57) was raised against the Daudi cell line. This antibody was found to interact with a 30-kDa protein on these cells and had agonistic properties. It enhanced the B lymphoblastoid accessory cell and interleukin 1 (IL-1)-dependent stimulation of PHA-treated murine thymocytes. The stimulatory effect of I57 on PHA-treated thymocytes was more pronounced at high, supraoptimal concentrations of the lectin. This was in contrast with the effect of IL-1 that failed to stimulate these cells treated with PHA at high concentrations. I57 also enhanced stimulation of thymocytes treated with IL-2 alone or with both PHA and IL-2. I57 exhibited by itself mitogenic activity for human T cells. These cells, treated with IL-2, were further stimulated by I57. I57 seems to be different from other agonistic antibodies that have been described so far.     

Lectins modulate the internalization of recombinant interferon-alpha A and the induction of 2',5'-oligo(A) synthetase

After binding to specific cell surface receptors, interferon-alpha (IFN-alpha) along with its receptor is internalized by the cells. However, the physiological significance of the internalization of IFN is not known. We have found that the lectin concanavalin A (ConA), which does not inhibit the binding of 125I-rIFN-alpha A, inhibits both the internalization of 125I-rIFN-alpha A and the rIFN-alpha A-induced increase in the levels of 2',5'-oligo(A) synthetase mRNA and enzymatic activity in the B lymphoblastoid cell line Daudi. The reduced level of IFN-induced 2',5'-oligo(A) synthetase in ConA-treated cells was due neither to direct inhibition of the enzymatic activity nor to generalized inhibition of protein or RNA synthesis. The dose-response curves were similar for the effect of ConA to inhibit 125I-rIFN-alpha A internalization and 2',5'-oligo(A) synthetase induction. The correlation between the ConA-mediated inhibition of both 125I-rIFN-alpha A internalization and 2',5'-oligo(A) synthetase induction suggests that internalization of rIFN-alpha A plays a role in the responses to rIFN-alpha A. However, since ConA inhibits protein mobility in the plasma membrane, it is possible that ConA is also preventing aggregation of IFN receptors or interactions between IFN receptors and signal transducing proteins in the plasma membrane that may be necessary for responses to IFN.     

Rabbit antiserum against Daudi-cell membrane fractions detects cell surface antigens present on subpopulations of human lymphocytes.

An antiserum was produced by immunization of rabbits with membrane preparations of a human lymphoblastoid B cell line, Daudi. After absorption with a human endometrial carcinoma cell line, this antiserum appeared to be specific for antigen(s) present on adult and fetal thymocytes as well as on tonsillar lymphocytes but absent, or present in very small amounts, on normal or phytohemagglutinin- (PHA) stimulated peripheral blood lymphocytes (PBL). When T and B cell-enriched fractions from tonsillar lymphocytes were tested with the anti-Daudi serum, the reactivity was equally distributed in each population. Among 13 human lymphoblastoid cell lines tested, reactivity was demonstrated on three out of four T cell lines, and on four out of nine B cell lines. The positive reacting B cell lines were derived from two African and two American Burkitt lymphomas. The antigen(s) described does not seem to be related either to human Ia-type antigens or to Epstein-Barr virus-associated antigens because these antigens are not present on fetal or adult thymocytes. Reciprocal absorption experiments indicate that this anti-Daudi serum detects the same antigenic structures present on certain subpopulations of T and B lymphocytes.