Mouse fibroblast (type I) and immune (type II) interferons: pronounced differences in affinity for gangliosides and in antiviral and antigrowth effects on mouse leukemia L-1210R cells

Anti-ganglioside anti-idiotypic vaccination: more than molecular mimicry

Surgery, chemotherapy, and radiation therapy are standard modalities for cancer treatment, but the effectiveness of these treatments has reached a plateau. Thus, other strategies are being explored to combine with the current treatment paradigms in order to reach better clinical results. One of these approaches is the active immunotherapy based on the induction of anti-tumor responses by anti-idiotypic vaccination. This approach arose from Jerne’s idiotypic network theory, which postulates that B lymphocytes forms a functional network, with a role in the establishment of the immune repertoires, in the regulation of natural antibody production and even in the establishment of natural tolerance. Due to the large potential diversity of the immunoglobulin variable regions, the idiotypes repertoire can mimic the universe of self and foreign epitopes, even those of non-protein nature, like gangliosides. Gangliosides are sialic acid-containing glycolipids that have been considered attractive targets for cancer immunotherapy, based on the qualitative and quantitative changes they suffer during malignant transformation and due to their importance for tumor biology. Although any idiotype could be able to mimic any antigen, only those related to antigens involved in functions relevant for organism homeostasis, and that in consequence has been fixed by evolution, would be able not only to mimic, but also to activate the idiotypic cascades related with the nominal antigen. The present review updates the results, failures and hopes, obtained with ganglioside mimicking anti-idiotypic antibodies and presents evidences of the existence of a natural response against gangliosides, suggesting that these glycolipids could be idiotypically relevant antigens.

The effect on MHC class II expression and apoptosis in placenta by IFNgamma administration

To investigate the possible mechanisms by which interferon gamma (IFNgamma) affects pregnancy, the expression of class II major histocompatibility complex (MHC) antigens and cleavage of DNA that is a hallmark of apoptosis in the placenta were examined by molecular biochemical techniques, and progesterone levels were examined by radioimmunoassay. The semi-quantitative analysis with reverse transcription-polymerase chain reaction showed that the expression of MHC class II antigen in placenta increased when rabbits were treated with high doses of IFNgamma compared with the control. However, immunohistochemical study suggested that IFNgamma did not affect MHC class II expression in trophoblasts, but had a stimulatory effect on its expression in maternal decidua and placental lymphocytes. DNA fragmentation analysis and terminal deoxynucleotidyl mediated-deoxyuridine triphosphate nick end labeling (TUNEL) assay indicated that the cleavage of DNA was detected in the placenta in both normal and IFNgamma-treated pregnancy. Quantitative analysis of apoptotic cells revealed an increase in trophoblasts treated with IFNgamma compared to those in normal pregnancy. Moreover, progesterone, which plays an important role in pregnancy, was reduced significantly in rabbits treated with IFNgamma. The results suggested that IFNgamma exerted its deleterious effect on pregnancy by inducing apoptosis in trophoblasts and by reducing the production of progesterone.

Simultaneous binding of mouse monoclonal antibody and streptavidin to heterobifunctional dendritic L-lysine core bearing T-antigen tumor marker and biotin

Thiolated T-antigen [Galbeta-(1-3)-GalNAcalpha, T-Ag] (6), derived in situ from thioacetate 5 was coupled to N-chloroacetylated glycylglycyl L-lysine dendritic cores (7-9) using high yielding substitution reactions to afford di- (10), tetra- (11), and octa-valent (12) glycodendrimers in good yields (76-86%). Heterobifunctional conjugate 14 was prepared as a biosensor from tetravalent conjugate 11 and biotin hydrazide 13 using TBTU strategy. In a solid-phase double sandwich enzyme linked immunosorbent assays (ELISA), biotinylated conjugate 14 was shown to bind to streptavidin used as a coating material. Mouse monoclonal anti T-Ag antibody (IgG3) and horseradish peroxydase-labeled goat anti mouse IgG, used for quantification, were found to bind T-Ag tetramer 14 immobilized on the surface of the streptavin layer. A typical saturation curve was observed for 14 while non-biotinylated tetramer 11 showed no binding in the entire concentration range. These results demonstrate the availability of both haptens toward the T-Ag antibody and streptavidin receptors.

Interferon-induced class II expression at the spongiotrophoblastic zone of the murine placenta is linked to fetal rejection and developmental abnormalities

Type II interferon is known to induce a plethora of gene expression involved in the humoral and cellular immunity. One of the multiple sites of action of gamma-IFN is the fetoplacental unit, where its role has not yet been clearly defined. We have previously shown in vitro that gamma-IFN may induce expression of class II MHC antigens on the spongiotrophoblast layer of the murine placenta, which under physiological conditions is negative for these antigens. Indeed, the absence of class II antigens from the placenta could be part of a mechanism evoked by fetal tissues to escape a host vs. graft reaction. In the present study we show that intraperitoneal in vivo administration of low doses of recombinant gamma-IFN to pregnant females specifically induces class II antigens on the spongiotrophoblast zone, increases fetal abortion, causes retardation of eye development in the fetuses and decreases fetal weight. This treatment also affects the maternal pathology as we witness a prominent hypersplenism in the mother accompanied by low levels of haematocrit, elevated IgG production and decreased granulocytic and thrombocytic counts. These results are specifically linked to the pregnant state of the mother, since virgin females do not develop any of the above abnormalities. Our results not only point to a new dimension in gamma-IFN's role during pregnancy, but may be of clinical importance for prophylaxis since administration of gamma-IFN to a pregnant female may lead to abortion, fetal abnormalities or cause haematologic disorders to the mother.

Conformational analysis of sialyloligosaccharides

The conformational properties of several sialyloligosaccharides present as terminal sequences in N- and O-linked carbohydrate groups of glycoproteins, have been analyzed based on the n.m.r. data of selected sialosides. The compounds examined include representatives of the alpha-D-NeuAc-(2----6)-beta-D-Gal-(1----4)-beta-D-GlcNAc, alpha-D-NeuAc-(2----3)-beta-D-Gal-(1----4)-beta-D-GlcNAc, alpha-D-NeuAc-(2----3)-beta-D-Gal-(1----3)-beta-D-GlcNAc, and alpha-D-NeuAc-(2----3)-beta-D-Gal-(1----3)-beta-D-GalNAc series. Two deuterated sialosides were prepared by enzymic sialylation of 6-deuterated galactose derivatives of methyl beta-D-galactopyranoside and lactoside. These were useful for the unambiguous establishment of the "gt" orientation of the flexible C-6 methylene unit of the galactose through 1H-1H coupling constants. Of all the (2----6) sialosides examined, only the deuterated di- and tri-saccharide afforded useful nuclear Overhauser enhancement data that could be used to evaluate the global minimum-energy conformations. Hard-sphere exoanomeric effect calculations estimated the glycosidic torsion angles for the global minimum-energy conformer of alpha-D-NeuAc-(2----6)-beta-D-Gal linkages to be -163/-132/61 degrees (theta, psi, and omega, respectively). However, the potential energy well surrounding this global minimum was very shallow and indicated a broad population distribution of conformers. These are illustrated by the isoenergy contour maps. The observation of n.O.e. between the H-3ax and H-6R of the galactose in two deuterated (2----6) sialosides, indeed supported the presence of one of the global minimum-energy conformers. The conformational analysis carried out for the di- and trisaccharide [alpha-D-NeuAc-(2----6)-beta-D-Gal-OMe and alpha-D-NeuAc-(2----6)-beta-D-Gal-(1----4)-beta-D-Glc-OMe respectively] was then extended to sialoside linkages of other tri- and penta-saccharides by comparison of their 1H- and 13C-n.m.r. chemical shifts. HSEA calculations for the (2----3) sialosides indicated the potential energy well containing the global minimum energy-conformer (theta, psi = -160 +/- 4, -11 +/- 2 degrees) was deeper than the one estimated for the (2----6) sialosides. The n.O.e. data are consistent with the distribution of the majority of conformers around the lowest-energy one in solution. CPK models highlighting the topographical differences between the lowest-energy conformations of alpha-(2----6) and alpha-(2----3) sialosides are presented.

Growth inhibitory effects of interferon-beta but not interferon-alpha on human glioma cells: correlation of receptor binding, 2',5'-oligoadenylate synthetase and protein kinase activity

The antiproliferative effects of human recombinant interferon-alpha (rIFN-alpha A) and interferon-beta (rIFN-beta ser) were assessed in vitro against seven human glioma cell lines. Further analysis of one of these lines (EFC-2) in response to rIFN-alpha A demonstrated a minimum growth inhibition by day 6 of treatment, whereas a 50% inhibition of cell growth was observed with a dose of 50 U/ml of IFN-beta ser. No significant growth inhibition was seen by rIFN-alpha A at doses up to 500 U/ml. Addition of rIFN-alpha A to rIFN-beta ser-treated EFC-2 cells neither suppressed nor augmented the antiproliferative response to IFN-beta ser. The binding of 125I-labeled rIFN-alpha A or 125I-labeled rIFN-beta ser to EFC-2 cells was inhibited competitively by increasing concentrations of either unlabeled rIFN-alpha A or rIFN-beta ser. This suggests that the cellular receptors for both rIFN-alpha A and rIFN-beta ser appear to be intact and appear to bind both agents equally. Furthermore, incubation of EFC-2 cells for 72 h with either rIFN-alpha A or rIFN-beta ser resulted in an increase in 2',5'-oligoadenylate (2-5A) synthetase activity 5-fold with rIFN-alpha A and 50-fold with rIFN-beta ser. Similarly, the 68-kD IFN-induced protein kinase was induced substantially with rIFN-beta ser but only slightly induced with rIFN-alpha A treatment. These results suggest that EFC-2 human glioma cells demonstrate a differential sensitivity in terms of growth inhibition to rIFN-beta ser and to rIFN-alpha A which appears to correlate with a differential induction of both intracellular 2-5A synthetase and protein kinase activity. These results cannot be explained solely on the basis of surface receptor binding of rIFN-alpha A and rIFN-beta ser. These data do suggest that, for human glioma cells in culture, type I IFN receptors may display a subtle architectural variation that allows equivalent binding of both IFN-alpha and IFN-beta ser, but allows an enhanced signal transduction and biological effect only after binding a specific IFN subtype.

Biologic response modifiers and urologic malignancies

Biologic response modifiers represent a new generation of immunotherapeutic agents. Although biologic response modifiers such as interferon have been moderately successful in treating B-cell malignancies, they have not been as successful for the management of advanced genitourinary malignancies. The measurement of biologic response modifiers in the urine may prove useful for directing intravesical BCG therapy for the treatment of superficial bladder cancer or for directing the use of prophylactic antibiotics for the management of urinary tract infections.

Characterization of L1210 S cells with low sensitivity to mouse interferon-gamma

A mouse leukemic cell line L1210 Sg with a low sensitivity to interferon-gamma (IFN-gamma) was described. On the nature of the antiviral action and binding of IFN, L1210 Sg cells were compared with L1210 m cell line which is sensitive to IFN-gamma. For a half reduction of the vesicular stomatitis virus-RNA synthesis, L1210 Sg cells required 500-fold more IFN-gamma than L1210 m cells did. However, both cell lines were induced to the antiviral state to the same extent with IFN-alpha or -beta. L1210 Sg and L1210 m cells were sensitive to the anti-proliferative action of IFN-alpha and -beta, but insensitive to IFN-gamma. (2'-5')Oligoadenylate synthetase was induced in these cell lines by IFN-beta, but not by IFN-gamma, which suggests that the induction of this synthetase may not be responsible for the antiviral action of IFN-gamma. No substantial difference between L1210 Sg and L1210 m cells was found in IFN receptors for IFN-gamma and IFN-beta either in number per cell or in their affinity to corresponding IFN type. However, differences were noted in time course profiles of cell-associated IFN-gamma at 37 C: in L1210 m cells, a rise-and-decay profile of IFN-gamma bound to cells was observed at 37 C, but in L1210 Sg cells, rise and slight decay was observed. On the other hand, a similar rise-and-decay curve of IFN-beta bound to these cells was observed. These results indicated that the low sensitivity of L1210 Sg cells to IFN-gamma may be due to this slight decay of receptor-bound IFN-gamma.

Differential effects of beta- and gamma-interferons on natural killer cell-mediated lysis of lung carcinoma cells

Peripheral blood lymphocytes from healthy donors (PBL) poorly lyse lung carcinoma cell lines A-549, A-427 and SK- MES-1 when tested in a short-term chromium release assay. When PBL are preincubated with human beta-interferon (IFN-beta), these cell lines are lysed with an efficacy comparable to that of erythroleukemia K-562 cells, the standard targets used in natural killer cell assays. However, when PBL are preincubated with gamma-interferon (IFN-gamma) instead, lysis of the lung carcinoma lines is little augmented. Unlabeled lung carcinoma A-549 cells block chromium release from labeled K-562 cells with non-boosted and IFN-gamma or IFN-beta-boosted effector cells. Also with the IFN-beta treated effectors, chromium release from A-549 targets is inhibited by unlabeled K-562 cells. Therefore, cells that lyse K-562 cells must be able to recognize A-549 cells, and, in the case of IFN-beta pretreated effectors, cause the killing of these cells as well. Data obtained with effector cells separated on discontinuous Percoll gradients also indicate that the same cells that lyse A-549 cells are responsible for lysis of K-562 cells. We conclude that in response to IFN-beta, effector cells previously able to lyse K-562, but unable to lyse A-549 targets, mature into fully competent killer cells capable of lysing tumor cells from lymphoid as well as from lung cancer origin. This effect is not elicited by IFN-gamma, indicating that killer cells respond differently to both interferon types.

An immunosuppressive factor interacts with membrane glycolipid asialo-GM1 structure

An immunosuppressive factor was isolated from sheep-erythrocyte hyperimmunized mouse serum. Possibilities of the adsorption of the immunosuppressive factor with glycolipid asialo-GM1 incorporated liposome (dipalmitoyl lecithin) and the blockade of immunosuppressive activity by specific anti-asialo-GM1 incorporated antibody were tested. We report here that: The immunosuppressive activity of the immunosuppressive factor decreased quantitatively if the factor was preincubated with asialo-GM1 incorporated liposome (GALP) as compared with that preincubated with liposome (LP) alone; and the presence of specific anti-asialo-GM1 antibody is able to protect in vitro anti-sheep erythrocyte (SRBC) response from the suppressive effect of the immunosuppressive factor in a dose-dependent fashion under similar cell culture condition. The results indicate that asialo-GM1 may be related to the receptor of immunosuppressive factor.

Inhibition of lipopolysaccharide-induced monocyte interleukin 1 secretion by gangliosides

Gangliosides known to be potent immunosuppressors are shown to inhibit the secretion of interleukin 1 (IL 1) by human monocytes stimulated with lipopolysaccharide (LPS). The inhibitory activity was observed either with mono-, di- or trisialogangliosides. The inhibition of LPS-induced IL 1 secretion was obtained in the presence of indomethacin, indicating that the ganglioside inhibitory capacity was not due to prostaglandin induction. Indeed the inhibitory activity seems to be mediated by an interaction with LPS molecules, preventing their ability to deliver a signal to monocytes for IL 1 secretion.

Sensitivity to alpha/beta-interferon is independent of N-linked complex-type oligosaccharides on cell-surface-membrane glycoproteins

The extent of involvement of carbohydrate structures in the mechanism of action of alpha and beta-interferon (IFN-alpha, IFN-beta) is undefined. In this report we examine the role of complex-type N-linked oligosaccharides in the response to these interferons. The response of mouse leukemia L 1210S cells, grown in the presence of swainsonine, an inhibitor of Golgi mannosidase II [Tulsiani, D. R. P., Harris, T. M. and Touster, O. (1982) J. Biol. Chem. 257, 7936-7939; Elbein A. D., Solf, R., Dorling, P. R. and Vosbeck, K. (1981) Proc. Natl Acad. Sci. USA 78, 7393-7397], to mouse IFN-alpha/beta, both with respect to antiviral and antigrowth effects, remains intact in spite of the total absence of complex-type N-linked oligosaccharides. Also, there is no difference in the response to human IFN-beta of a parental Chinese hamster ovary cell line and a mutant lacking beta-N-acetylglucosaminyltransferase I and therefore unable to synthesize complex-type N-linked oligosaccharides [Stanley, P., Callibot, V. and Siminovitch, L. (1975) Cell 6, 121-128]. These results are significant in permitting the conclusion that the carbohydrate-specific binding of IFN-alpha and IFN-beta to gangliosides cannot be due to a similarity of the ganglioside carbohydrate to that of a glycoprotein containing a complex-type N-liked oligosaccharide.

Cloning and expression of the chromosomal immune interferon gene of the rat

The chromosomal immune interferon gene of the rat (IFN-gamma) was identified by screening a recombinant rat lambda phage library with a human IFN-gamma cDNA probe. In contrast to the genes of other rat IFNs, this rat IFN-gamma chromosomal gene contains introns and its structural organization closely resembles that of the human and murine IFN-gamma genes. The rat IFN-gamma gene encodes a signal sequence of 19 amino acids followed by the mature IFN-gamma protein of 137 amino acids. The gene was expressed under control of the simian virus 40 (SV40) early promoter in Chinese hamster ovary (CHO) cells deficient in dihydrofolate reductase (DHFR) after co-transformation with a plasmid containing the mouse DHFR gene. Initial transformants with a DHFR+ phenotype produced IFN-gamma titres ranging from 20 to 1600 units/ml. After stepwise increases in the concentration of methotrexate (MTX) in the growth medium of transformed CHO cells, MTX-resistant clones producing 80 000-100 000 units per ml were isolated. Protein analysis of supernatants of these MTX-resistant cells by polyacrylamide gel electrophoresis revealed a product with an apparent mol. wt. of 18 000 daltons which was not detectable in the growth medium of DHFR+ transformants that did not produce IFN. The product was identified as rat IFN-gamma and constituted approximately 5% of the proteins excreted from these cells.

Potentiation of the antiviral and anticellular activities of interferons by mixtures of HuIFN-gamma and HuIFN-alpha or HuIFN-beta

Treatment of transformed human amnion WISH cells or human diploid fibroblasts (FS-4) or human fibroblasts trisomic for chromosome 21 (GM2767) with mixtures of human interferon gamma (HuIFN-gamma) and either natural leukocyte HuIFN-alpha or recombinant HuIFN alpha 2 or natural fibroblast HuIFN-beta resulted in potentiation of the antiviral activity of these IFNs. Pretreatment for 22 h of WISH cells with HuIFN-gamma followed by the addition of either HuIFN-alpha or HuIFN-beta resulted in significant potentiation of the antiviral action of these IFNs. The range of potentiation was 3 to 15-fold. Similar potentiation was observed when these IFNs were added simultaneously to the cells. Pretreatment for 22 h of WISH cells with either HuIFN-alpha or HuIFN-beta followed by the addition of HuIFN-gamma also resulted in significant increase of the antiviral protection against virus yield (3 to 39-fold). The level of the potentiation was higher in comparison with the antiviral activity observed when all these IFNs were added at the same time. Treatment of human FS-4 and GM2767 fibroblasts with HuIFN-gamma and either HuIFN-alpha or HuIFN-beta revealed potentiation of anticellular properties of these IFNs. The level of potentiation of anticellular activity was in the range of 2.7- to 9.4-fold. In the majority of the experiments, maximum potentiation of either antiviral or anti-cellular activity was observed when mixtures of equivalent concentrations of IFNs were used. The antiviral and anticellular functions of natural or recombinant IFN-alpha and fibroblast IFN-beta were potentiated usually to a similar degree by the presence of IFN-gamma. In contrast, combinations of HuIFN-alpha (natural or recombinant) and HuIFN-beta, in the absence of HuIFN-gamma, did not potentiate the anticellular or antiviral activity.

Roles of gangliosides in the surface properties of normal and malignant cells

Gangliosides are ubiquitous plasma membrane components whose structural characteristics make it possible to establish multiple interactions with the pericellular microenvironment. Several receptorial activities of gangliosides have already been recognized and the possible involvement of gangliosides in growth control, adhesion, differentiation, and immunologic recognition is at present under study. The study of the role of gangliosides in cell growth control has been approached in our laboratory by investigating whether the ganglioside changes found in SV40-transformed Balb/c3T3 cells (SV3T3 cells) revert to a normal pattern in a variant of SV3T3 cells with growth properties similar to those of normal parental cells, the concanavalin A-selected SV3T3 revertant cells. These latter cells showed the same reduction of the more complex gangliosides as found in SV3T3 cells, while their amount of II3NeuAC-LacCer (GM3) was greatly increased compared to that found in either normal or transformed 3T3 cells. In order to study the role of gangliosides in the adhesion process, we analyzed the ganglioside structure of the so-called substrate-attached material, a cell surface structure involved in the adhesion of cells to substrate, from cultures of Balb/c3T3, SV3T3, and concanavalin A-selected SV3T3 revertant cells and from cultures of a system of transformed cells with different metastatic potential: the nonmetastatic B77-3T3 and the highly metastatic AA6 cells. Compared to normal cells, all the transformed cells contained smaller quantities of gangliosides in their substrate-attached material.(ABSTRACT TRUNCATED AT 250 WORDS)

Synergistic antiviral and antiproliferative activities of Escherichia coli-derived human alpha, beta, and gamma interferons

The antiviral and antiproliferative effects of highly purified Escherichia coli-derived human interferons (IFNs) were examined in human melanoma cells (Hs294T). Antiproliferative activity was monitored by measuring inhibition of cell multiplication, and antiviral activity was determined by inhibition of herpes simplex virus type 1 replication. Treatment of cells with IFN-gamma in combination with IFN-alpha A or IFN-beta 1 resulted in potentiation of both antiproliferative and antiviral activities. In contrast, combination treatments composed of IFN-alpha A and IFN beta 1 yielded inconsistent results. Some combinations reflected additive responses, whereas others were antagonistic. To examine correlations between IFN-induced biological activities and interactions of the different IFNs with cell surface receptors, in vivo [35S]methionine-labeled IFN-alpha A was prepared. Binding studies indicated the presence of 2,980 +/- 170 receptors per cell, each with an apparent Kd of (8.4 +/- 1.3) X 10(-11) M. Results from competitive binding studies suggested that Hs294T cells possess at least two types of IFN receptors: one which binds IFN-alpha A and IFN-beta 1 and another to which IFN-gamma binds.

Isolation and characterization of an interferon-resistant cell line deficient in the induction of (2'-5')oligoadenylate synthetase activity

To screen for cells with different sensitivities to interferon (IFN), NIH 3T3 mouse fibroblasts were subcloned and examined for their response to IFN treatment. Of 30 clones tested, 2 appeared to be relatively resistant to IFN, since the replication of both vesicular stomatitis virus and mengovirus was not inhibited, even in the presence of 1,000 U of IFN per ml. One resistant (A10) and one sensitive (A5) clone were further analyzed. In both clones, murine leukemia virus replication was equally inhibited by IFN, indicating the presence of functional receptors for IFN in the resistant clone. Using the (2'-5')oligoadenylate (2-5A) radiobinding assay, we could demonstrate that both clones contained the RNase L protein. Furthermore, this enzyme appears to be active, since a similar reduction in the rate of protein synthesis was evident after the introduction of exogenous 2-5A to the cells. We also analyzed the activity of another enzyme in the 2-5A pathway, namely, 2-5A synthetase. In the sensitive cells (A5), the induction of enzyme activity was proportional to the IFN concentration used, reaching a maximum of more than a 10-fold increase over the background of untreated cells. However, little if any induction over the basal activity was observed in the resistant cells (A10) when similar doses of IFN were used. It is thus probable that the lack of induction of 2-5A synthetase activity by IFN in A10 cells is at least partly responsible for their relative resistance to IFN treatment.

Requirement for IFN gamma in potentiation of interferon's antiviral and anticellular activities: identity of mouse and human systems

Potentiation was originally demonstrated as a nonadditive, synergistic enhancement of interferon (IFN) activity in the mouse system for mixed preparations containing MuIFN-alpha/beta and MuIFN-gamma. Potentiation of the antiviral and anticellular activities has now been studied for mouse and human systems, and in both systems IFN-alpha and IFN-beta interacted synergistically with IFN-gamma, but not with each other. Further, the antiviral and anticellular activities of IFN-alpha and IFN-beta were potentiated equally by IFN-gamma. Potentiation was demonstrated for HuIFNs with specific activities of 10(7) U/mg of protein and higher. Naturally produced and recombinant HuIFN-alpha s had the same relative abilities to be potentiated by HuIFN-gamma. It was concluded that IFN-gamma with either IFN-alpha or IFN-beta was essential for potentiation, that potentiation of anticellular and antiviral actions occurred in similar manners, and that a close correlation existed between potentiation in mouse and human systems. These results suggest that potentiation was caused by the interaction of two dissimilar IFN types (immune versus virus-type) and that potentiation studies in the mouse may be directly relevant for humans.

Studies on purification of human gamma interferon: chromatographic behavior of accompanying IL2 and B-cell helper activity

Gamma interferon (gamma IFN) was produced in human lymphocyte cultures stimulated by PHA. Titers were in the range of 10,000-30,000 U/ml. Crude gamma IFN was adsorbed on silicic acid, from which the antiviral activity was eluted by a buffer containing a high salt concentration and ethylene glycol. This treatment allowed quantitative recovery of gamma IFN with a specific activity of 5 X 10(5)-1 X 10(6) U/mg of proteins. IL2 and B-cell helper activities were adsorbed and eluted from silicic acid together with the antiviral activity. This finding might be of practical interest for the purification of these lymphokines, particularly IL2. Gamma interferon was further purified on Blue Sepharose to a specific activity of 2 X 10(7) U/mg. The resulting preparations still contained IL2 and B cell helper activities. However, taking advantage of the differences in apparent hydrophobicity and in isoelectric point, we were able to dissociate antiviral activity from lymphokines. Such dissociation should facilitate the study of the biological properties of human natural gamma IFN.

Factors acting upon differentiation of mouse teratoma in vivo: spontaneous differentiation and effects of interferon

CE/44 mouse teratocarcinoma and its sublines isolated at the Institut Pasteur were studied comparatively in 129/Sv, hybrid F1 129 X NCS, and NCS strains of male and female mice with respect to the age of the animals and the type of the tumor graft (ascitic or subcutaneous solid tumor). The tumors grow more quickly and with more malignant characteristics in female mice. The tumor "takes" and growth are equally good in the hybrid F1 as in 129/Sv mice, but its maintenance in hybrid F1 is not so stable, for a tumor loss and a lessening of differentiation could be seen through the passages in this strain. In the NCS strain the tumor grows very rarely. Interferon slowed the subcutaneous tumor growth and favored neuroepithelial tissue differentiation, but allowed selective growth of one cellular type, the "trophoblastic" one, which is very malignant, and which, in the ascitic tumor, caused death in female animals by a hemorrhagic syndrome.

Isolation and characterization of an interferon-resistant cell line deficient in the induction of (2'-5')oligoadenylate synthetase activity

To screen for cells with different sensitivities to interferon (IFN), NIH 3T3 mouse fibroblasts were subcloned and examined for their response to IFN treatment. Of 30 clones tested, 2 appeared to be relatively resistant to IFN, since the replication of both vesicular stomatitis virus and mengovirus was not inhibited, even in the presence of 1,000 U of IFN per ml. One resistant (A10) and one sensitive (A5) clone were further analyzed. In both clones, murine leukemia virus replication was equally inhibited by IFN, indicating the presence of functional receptors for IFN in the resistant clone. Using the (2'-5')oligoadenylate (2-5A) radiobinding assay, we could demonstrate that both clones contained the RNase L protein. Furthermore, this enzyme appears to be active, since a similar reduction in the rate of protein synthesis was evident after the introduction of exogenous 2-5A to the cells. We also analyzed the activity of another enzyme in the 2-5A pathway, namely, 2-5A synthetase. In the sensitive cells (A5), the induction of enzyme activity was proportional to the IFN concentration used, reaching a maximum of more than a 10-fold increase over the background of untreated cells. However, little if any induction over the basal activity was observed in the resistant cells (A10) when similar doses of IFN were used. It is thus probable that the lack of induction of 2-5A synthetase activity by IFN in A10 cells is at least partly responsible for their relative resistance to IFN treatment.

Action of different gamma interferons on mouse leukemic cells resistant to alpha and beta interferons

Phytohemagglutinin-induced interferon-gamma (IFN-gamma) was reported to act on mouse leukemic L1210 R cells resistant to IFN-alpha and -beta. Results reported here show that these cells are also sensitive to various preparations of murine IFN-gamma derived from different sources and purified to different degrees and that lymphokines present in the preparations are not involved in the antiviral effect of IFN. In addition, IFN-gamma preparation increases concanavalin A binding to L1210 S and L1210 R cells indicating that the sensitivity of L1210 R cells to IFN-gamma is not limited to its antiviral effect.

Interferon-induced increase in neuraminidase-releasable sialic acid and glycosphingolipid metabolism in mouse lymphoma and L1210 leukemic cell lines: correlation with susceptibility to natural killer cell-mediated lysis

Changes in sialoglycoconjugates and glycosphingolipid (GSL)5 metabolism were demonstrated in mouse EL4, P52 and YAC-1 lymphoma and L1210 leukemia cell lines treated with beta-interferon (IFN). Expression of cell surface (neuraminidase-releasable) sialic acid on IFN-treated cells was markedly elevated (three- to six-fold). The increase in neuraminidase-releasable sialic acid is contributed by sialoglycoproteins and particularly by cell-surface gangliosides in IFN-treated cells. Incorporation of [3H]-galactose into all GSL was elevated in IFN-treated cells. Thin-layer chromatographic analysis of GSL of IFN-treated cells showed an increase in several GSL homologues with striking changes in ganglioside with chromatographic migration of GM2, GM1, and GD1a relative to control cells. IFN-treated tumor-cell lines became resistant to lysis by virus-induced IFN-activated natural killer (NK) cells, as shown previously, but addition of neuraminidase to IFN-treated and untreated cells caused only a moderate increase in NK-sensitivity. This suggests that IFN-mediated protection of target cells from NK lysis was not due to a preferential masking of target structure by elevated levels of sialic acid. These membrane-associated changes in GSL and sialic acid in IFN-treated cells may be potentially significant, because a correlation between certain GSL expression, sialic acid phenotype and susceptibility of target cells to NK-cell-mediated lysis have been found in several other systems.

Ultrastructural distribution of interferon receptor sites on mouse L fibroblasts grown in suspension: ganglioside blockade of ligand binding

Murine beta-interferon (IFN) receptors on L929 cells grown in suspension culture were visualized by indirect immunoferritin electron microscopy. Ferritin label on these cells was associated primarily with the coated areas and coated pits of the membrane, in contrast to previous observations with L929 cells grown in a monolayer, which did not reveal such coated areas or pits but showed ferritin label distributed randomly on the cell membrane (Kushnaryov et al., Infect. Immun. 36:811-821, 1982). On about 15% of the cell sections from suspension-grown cells, the ferritin label was found outside coated membrane areas. These findings suggest that different cell populations exist with respect to the localization and possibly the affinity of IFN receptors. In the same experiment, exogenously added gangliosides blocked the binding to cell surfaces not only of 125I-labeled IFN but also of unlabeled IFN as revealed by an immunospecific ferritin labeling technique, providing direct evidence that gangliosides interfere with the binding of IFN to specific receptor sites on the surface of mouse L929 cells. These studies establish that the binding of IFN to cell membranes, depending on cell growth conditions, can involve coated areas and coated pits, to which certain hormones and toxins have been shown to bind.

A unique set of polypeptides is induced by gamma interferon in addition to those induced in common with alpha and beta interferons

Human immune interferon (IFN-gamma) differs from leukocyte interferon (IFN-alpha) and fibroblast interferon (IFN-beta) in cell origin, inducing agents, physical and biological properties and amino acid sequence. These differences have led to interest in possible differences in the biological properties of IFN-gamma compared with IFN-alpha and IFN-beta. IFN-gamma has the same broad range of biochemical and biological actions as do IFN-alpha and IFN-beta, although relative potencies vary depending on the cell type and function investigated. There has so far been no direct evidence that IFN-gamma alters normal cell functions differently from other interferons. We report here striking qualitative and quantitative differences in the intracellular response of human fibroblasts to IFN-gamma compared with IFN-alpha and IFN-beta. Two-dimensional gel electrophoresis demonstrates, in addition to the induction of a common group of polypeptides, the existence of a set of polypeptides whose synthesis is uniquely induced by IFN-gamma.

Interferon potentiation occurs at a pretranscriptive stage

Combination of interferon-gamma (IFN-gamma) with a mixture of IFN-alpha and IFN-beta (IFN-alpha/beta) produce a potentiated antiviral state. The onset of induction of the potentiated antiviral state was observed to be more rapid than the onset of induction of either IFN-gamma or IFN-alpha/beta separately. The onset of viral resistance to combined IFN-gamma and IFN-alpha/beta (25 U/ml each) began after 2 h, compared with 4 h for IFN-alpha/beta (25 U/ml) alone and 8 h for IFN-gamma (25 U/ml) alone. By adding actinomycin D at various times after interferon treatment, this more rapid onset of induction of the potentiated antiviral state was shown to be the result of a more rapid onset of cellular transcription. To determine whether increased concentrations of IFN-alpha/beta acted through a similar mechanism, we performed kinetic experiments with several concentrations of IFN-alpha/beta and with combined preparations of IFN-gamma and IFN-alpha/beta. The findings showed that treatment of cells with the combined interferons and with a high concentration of IFN-alpha/beta resulted in more rapid onset of cellular transcription and more rapid development of the antiviral state. Thus, the cellular response to potentiation appeared to involve perception by the cell of the combined interferons as a higher-than-expected effective concentration of interferon, resulting in more rapid onset of cellular transcription of mRNA for antiviral proteins.

Grafting of triggering site onto lymphocytes; distribution of grafted dinitrophenyl groups on cell surface glycoproteins and glycolipids

Quantitative investigation of membrane-bound sialoglycoconjugates on lymphocyte surface was performed by chemical modification of the sialic acid residues with radioactive N4-dinitrophenyl-L-2,4-diaminobutyric acid hydrazide (DNP-DABH). This labeled both glycoproteins and glycolipids with concomitant preservation of the mitogenic activity by multivalent hapten binding protein (anti-DNP antibody). Under conditions where maximum stimulation of thymocytes occurred radioactive DNP-DABH labeled 1.1 X 10(7) glycolipids molecules/cell but, only 3 X 10(6) glycoproteins molecules/cell. When B lymphocytes, which do not undergo DNP-mediated stimulation were used, glycolipids labeling could not be detected. Major differences between stimulation committed and non-committed DNP-modified lymphocytes was the amount of ligand attached to the cell surface sialoglycolipids (gangliosides).

The antiviral action of interferon

On interferon treatment cells develop an antiviral state. This requires time and RNA and protein synthesis. At least six polypeptides and two enzymes have been reported to be synthesized in increased amounts in response to interferon and a multiplicity of effects have been attributed to it. Interferon has been reported to inhibit virus growth at the level of the uncoating of the virus, virus RNA and protein synthesis and virus maturation. This has led to the acceptance of a multisite model for interferon action. The evidence for this and for the role of two known interferon-mediated enzymes, the 2-5A synthetase and protein kinase, are reviewed.

Effects of human gamma interferon on cell growth, replication of virus and induction of 2'-5'oligoadenylate synthetase in three human lymphoblastoid cell lines and K562 cells

Human alpha interferon (IFN-alpha) and beta interferon (IFN-beta) showed antiviral and anticellular effects on human lymphoblastoid Daudi and P3HR-1 cells, but up to 1,000 units/ ml of gamma interferon (IFN-gamma) showed no such effect. Though a fairly high level of dsRNA-dependent 2'-5'-oligoadenylate synthetase (2-5A synthetase) was found in Daudi cells, treatment of these cells with IFN-alpha and beta enhanced the enzyme level in cells at least four-fold, but IFN-gamma did not show any such effect. Lymphoblastoid Raji cells were insensitive to the antiviral and anticellular activities of IFN-alpha, beta and gamma, but 2-5A synthetase was induced in cells by the treatment with IFN-alpha and beta, though the enzyme level was lower than that found in interferon-treated Daudi cells. Human leukemic K562 cells were completely insensitive to IFN-alpha, beta and at the same time to IFN-gamma with regard to the antiviral, anticellular activities and to the induction of 2-5A synthetase.

Inhibition of mouse fibroblast interferon by gangliosides. Differential effects on biological activity and on induction of (2'--5')oligoadenylate synthetase

Gangliosides are potent inhibitors of the antiviral activity of mouse fibroblasts and other beta-interferons. We have compared the effects of gangliosides on antiviral and antigrowth activities of mouse fibroblast interferon and on the induction of (2'--5')oligoadenylate synthetase, one of the enzymes implicated in the antiviral state induced by interferon. Whereas both biological effects appear to be inhibited by gangliosides in an analogous fashion, inhibition of induction of (2'--5')oligoadenylate synthetase does not correlate with inhibition of vesicular stomatitis virus replication. Ganglioside concentrations that inhibit the interferon-induced (2'--5')oligoadenylate synthetase to levels close to those of uninduced cells, still allow for a 100--1000-fold reduction of viral yield. Significantly higher ganglioside concentrations are required to prevent completely the antiviral effect. This biphasic relationship between (2'--5')oligoadenylate synthetase levels and inhibition of viral yield suggests that no or very small increases in synthetase levels are involved in inhibition of virus by between two and three orders of magnitude.

Molecular cloning of human immune interferon cDNA and its expression in eukaryotic cells

Starting with mRNA derived from Staphylococcal enterotoxin A induced human splenocytes, dsDNA was synthesized and inserted into unique BamHI site of the eukaryotic expression vector pSV529 (1). A recombinant plasmid containing human immune interferon (IFN-gamma) cDNA was identified by hybridization of plasmid inserted DNA bound onto nitrocellulose filters with mRNA derived from SEA-induced splenocytes, translation of the eluted RNA in Xenopus laevis oocytes and assaying for IFN activity. Plasmids containing the entire human IFN-gamma cDNA sequence were identified by colony hybridization and were sequenced. A unique coding region was identified which predicted a protein of 166 amino acids, the 20 N-terminal amino acids of which presumably represent a signal peptide. After transfection of monkey cells with plasmid DNA isolated from one of the recombinant clones (pHIIF-SV-gamma 1), IFN was excreted into the culture medium. This IFN was not distinguishable from human IFN-gamma by serological criteria or by cell target species specificity.

Evidence for the identity of murine gamma interferon and macrophage activating factor

Macrophages activating factor (MAF) in mouse lymphokine preparations was quantitated using a tumor cell cytotoxicity assay. MAF activity was compared with gamma interferon (IFN-gamma) activity, and the lymphokine mixture subjected to a variety of protein fractionation procedures. No significant difference in the ratio of MAF activity to IFN activity was observed following any of the fractionation steps, even after MAF had been purified to a specific activity of 1 X 10(6) u/mg protein. Gel permeation using high pressure liquid chromatography showed a coincident peak of MAF and IFN activity at approximately 55 kD. Both activities were reduced at similar rates following heating at 56 degrees C or incubation at 4 degrees C in pH 2 buffer. Finally, induction of lymphokines using different inducers (mitogens or antigens) or cell populations always resulted in similar ratios of MAF activity to IFN activity. These results support the hypothesis that MAF and IFN-gamma are identical proteins.

Inhibition of beta-interferon augmentation of murine natural killer cytotoxicity by gangliosides

Interferons cause augmentation of natural killer (NK) cell activity, which might be a major reason for their antitumor effect. Antiviral and antigrowth effects of mouse beta interferon are inhibited by mono-, di-, and trisialogangliosides commonly found in brain extracts, but also in membranes of many other cells. Results presented in this report show that preincubation of mouse beta interferon with a brain ganglioside mixture or its isolated major components Gm1, Gd1a, Gd1b, and Gt1b (see Aberrations) prior to addition to effector spleen cells, inhibits NK-cell enhancement due to interferon in a dose-dependent manner. When spleen cells are treated with individual gangliosides alone, spontaneous NK cell activity is not affected. Pretreatment of effector cells with gangliosides prior to addition of interferon does not inhibit subsequent augmentation of NK cell activity by beta interferon. Also, target susceptibility remains unaltered in the presence of gangliosides. Thus the inhibitory effect of gangliosides appears to involve competition for interaction of beta interferon with the NK cells.