Studies of the mechanism by which gangliosides inhibit the proliferative response of murine splenocytes to concanavalin A

CELLULAR

IMMUNOLOGY

104,7 l-78 (1987)

Studies of the Mechanism by Which Gangliosides Inhibit the Proliferative Response of Murine Splenocytes to Concanavalin A’ DONALD M. MARCUS, ALICIA DUSTIF&* ISABELDIEGO, SUSANOSOVITZ, AND DOROTHY

E. LEWIS

Departments OfMedicine, Microbiology, and Immunology, Howard Hughes Medical Institute, Baylor College ofMedicine, One Baylor Plaza, Houston, Texas 77030 Received July 22, 1986; accepted August 24, 1986 Gangliosides are known to inhibit the proliferative response of murine and human lymphocytes to antigens and mitogens in vitro. In this study the response of murine spleen cells to concanavalin A (Con A) was used as a model system. Analysis of the cellular events by flow cytometry revealed that during the first 24 hr of culture the effect of gangliosides on Con A-treated cells was minimal. At 48 hr, however, more of the ganglioside-treated cells were in Go/G1, the cells contained more RNA, and fewer cells were in S phase. These data indicate that gangliosides inhibit the transition of the cells from Go/G, into the S phase of the cell cycle. Expression of the interleukin 2 (IL-2) receptor, as measured by the binding of a monoclonal antibody to the receptor, was not inhibited by the gangliosides. Binding of ‘z51-labeled recombinant IL-2 to cells cultured for 48 hr with Con A was inhibited by ganglioside GDlp but not by asialo GM,. Inhibition was much more effective if the gangliosides were preincubated with IL2 before addition of cells, but no inhibition was observed if the cells were preincubated with gangliosides and the unbound gangliosides were washed out prior to addition of the IL2. These data suggest that interference with the binding of IL-2 to the high-affinity IL-2 receptor of activated T lymphocytes plays an important role in the inhibition of Con A-induced proliferation. 0 1987 Academic Press., Inc.

INTRODUCTION The ability of gangliosides to suppress in vitro proliferative responses of mononuclear cells to antigens and mitogens has been documented in a number of publications (1-7, reviewed in 8). Studies of the mechanism of inhibition have revealed that Con A3-stimulated DNA and RNA synthesis were inhibited (l), and that gangliosides partially suppress the influx of calcium into Con A-stimulated murine lymphocytes (9). Prolonged preincubation of mononuclear cells with gangliosides was also shown to inhibit their function in in vitro proliferative responses (3-7). The ability of gangliosides to inhibit the IL-Zdependent proliferation of T-cell lines was noted in two re-

’ This work was supported by Research Grants AI 177 12 and RR-O5425 from the U.S. Public Health Service. Computational assistance was provided by the Clinfo Project, funded by the Division of Research Resources of the National Institutes of Health under Grant RR-00350. * Present address: Sigma Xi, 345 Whitney Ave., New Haven, Conn. 065 11. ’ Abbreviations used: Con A, concanavalin A, FCS, fetal calf serum, IL2, interleukin 2. 71 0008-8749187 $3.00 Copyright Q 1987 by Academic Press, Inc. All rigbts of rerwdwtioo in any form reserved.

72

MARCUS

ET AL.

TABLE 1 Structures of Glycosphingolipids” Asialo G,,,, GM,

G Dl# G Dlb Gr,b Sialosylparagloboside (SW

G~l-3GalNAc@l-4Galj31-4Glc@l-lCer Gal~I-3GalNAc~1-4[NeuAc~2-3]Gal~l-4Glcj3l-lCer NeuAccY2-3Gal@1-3GalNAcj31-4[NeuAc~2-3]Gal~l-4Glc~l-lCer Gal~l-3GalNAc~1-4[NeuAca2-8NeuAc~2-3]G~l-4Glc~l-lCer NeuAccu2-3Gal~1-3GalNAc~l~[NeuAccu2-8Ne~c~2-3]G~~l-4Glc~l-lCer NeuAcaZ-3Galj3 l -4GlcNAcj31-3Gala

1-4Glc@ I - 1Cer

’ Gal, Dgalactose; Glc, ~-glucose; GalNAc, N-acetyl-D-galactosamine; GlcNAc, N-acetyl-~glucosamine; NeuAc, IV-acetyl-neuraminic acid; Cer, ceramide (N-acylsphingosine).

cent reports (10, 1 I), and binding of IL-2 to a ganglioside column was also noted in the latter study. Several investigators have suggested that gangliosides exert their effect by interacting with cell membranes and interfering with the normal sequence of events involved in cellular activation. The studies reported below confirm the findings of Robb (12) that gangliosides interfere with the binding of IL-2 to high-affinity IL-2 receptors of activated lymphocytes. They extend these observations by demonstrating that the expression of the IL-2 receptor is not inhibited, and that gangliosides inhibit the transition of cells from Go/G, into S phase of the cell cycle. MATERIALS

AND METHODS

Gangliosides. Mixed bovine brain gangliosides and gangliosides GM,, Gnu,, Gma, GTlb (Table 1) were generously provided by Fidia Research Laboratories, (Abano Terme, Italy) and their purity was verified by thin-layer chromatographic analysis; Golb required additional purification by high-performance liquid chromatography. Asialo GM1 and sialosylparagloboside (SPG) were prepared as described previously (13). [3H]Thymidine (6.7 Ci/mM) and recombinant human ‘25I IL-2 (27.8 &i/mg) were obtained from New England Nuclear Research Products. Human recombinant IL-2 was obtained from Amgen Biologicals (Thousand Oaks, Calif.). Rat monoclonal antibody (7D4) directed against the murine IL-2 receptor was a gift from Dr. Ethan Shevach (14). Splenocyte cultures. Single spleen cell suspensions from 12- to 18-week-old normal BALB/cByJ female mice were prepared in sterile RPM1 1640 containing 2 &glummine, 0.02 MHepes, 100 U penicillin/ml, and 100 rg streptomycin/ml. Erythrocytes were lysed with tris ammonium chloride, and the spleen cells were washed three times with RPM1 and resuspended in RPM1 supplemented with 10% fetal calf serum (FCS). Cell viability was determined by trypan blue exclusion. Cells were aliquoted in 96-well Costar plates (Costar, Cambridge, Mass.) at a concentration of 2 X lo5 cells per well, and the total culture volume was 220 &well. Cultures were maintained in a humidified atmosphere of 8% CO*-92% air at 37°C. Con A (Pharmacia Fine Chemicals) was added at a concentration of 1 pg/ml, cells were cultured for 24 hr, 1 &i [3H]thymidine was added to each well, and the cells were harvested 18 hr later on a PhD cell harvester (Cambridge Technology, Inc., Cambridge, Mass.) using glass filter paper. Liquiscint scintillation cocktail (2 ml; National Diagnostics) were ali-

GANGLIOSIDE

INHIBITION

OF PROLIFERATION

73

quoted to each vial and the samples were counted in a scintillation counter. Solutions of pure gangliosides were prepared daily by evaporating stock solutions under N2 and dissolving the gangliosides in either sterile RPM1 or PBS without FCS and subjecting them to sonication in a water bath for several minutes; dilutions were made in RPM1 medium. Binding of ‘251-labeled IL-2 to Con A-stimulated splenocytes was performed by a modification of the method of Robb et al. (15). Splenocytes were cultured with 1 pg/ml of Con A as described above for 48 hr, washed, and incubated in medium with FCS twice for 30 min at 37°C to dissociate bound IL-2 from their membranes. After washing in RPM1 containing 10% FCS, 1 X lo6 cells were added to each tube, the suspension was centrifuged, and the supematant was removed. The cells were resuspended in 0.1 ml of ‘251-labeled IL-2 and incubated at 37°C for 30 min with gentle mixing. At the end of the incubation period 1 ml of cold complete medium was added, and the cells were washed three times in cold medium and counted. To test the effect of gangliosides on ‘*‘I-labeled IL-2 binding to the splenocytes, the IL-2 and gangliosides were mixed and preincubated at room temperature for 30 min before addition to the cell pellet. In other experiments, the mixture ofgangliosides and radiolabeled IL-2 was added directly to the cell pellet without preincubation. Binding of 704 to splenocytes. Splenic mononuclear cells were cultured for 48 hr in the presence of Con A (1 pg/ml) or Con A plus GDla (200 pg/ml) as described above. After washing with PBS the cells were incubated with an optimal amount of 7D4 for 30 min at 4°C washed, and incubated with fluorescein-labeled rat antibody against mouse kappa chains (Becton-Dickinson, Mountain View, Calif.). The cells were washed, fixed in 1% paraformaldehyde, and analyzed by flow cytometry. Flow cytometric analysis. A Coulter Epics V flow cytometer was used for forward angle light scatter, 90” light scatter, and fluorescence measurements. Single and dual parameter histograms were acquired and control samples were compared to test samples by simple integration of at least 20,000 events. The analysis of cellular DNA and RNA was done using modifications of the techniques of Braylan et al. ( 16) and Wallen et al. ( 17). Briefly, cells were initially fixed in 1% paraformaldehyde, and in order to allow the acridine orange to enter the cells they were fixed in ethanol by dropwise addition of an equal volume of 100% ethanol until the ratio of 1% paraformaldehyde:ethanol was 1: 1. The cells were fixed for 30 min and then centrifuged and resuspended in an equivalent volume of 15 pg/ml a&dine orange in solution A. Solution A was composed of 63.5 ml 0.2 MNaH2P04, 36.5 ml 0.1 M citric acid, 0.0372 g Na EDTA (1 mM), and 0.877 g NaCl(O.15 N) at a pH of 6.0. For simultaneous analysis of DNA and RNA, a&dine orange was exited by 488 mm at 500 mW. The fluorescent emissions were split using a 590-mm dichroic filter, and linear integrated red and green fluorescence signals were collected. The green fluorescence was collected using a 550-mm short pass filter after a 5 15-mm laser light-blocking filter. The red fluorescence was collected after a 630-mm long pass filter. Simultaneous measurement of 90” light scattering and green or red fluorescence was made using a 488-mm dichroic filter before the 5 15-mm interference filter into a separate 90” light scattering photomultiplier tube. The data in Table 2 were derived by integration of the 2dimensional histograms to obtain the percentage of cells in each DNA region ( 18). The number of cells on the Y axis (RNA) above a certain channel (the medium control had 10-l 5% of the events above that channel) was determined for each DNA region of the histogram.

74

MARCUS

ET AL.

TABLE 2 Effect of Glycolipids on Con A-Activated Splenocytes after 48 hr in Culture” Percentage cells Treatment

Go/G

S

Medium only Con A Con A + Gh(, Con A + Go,.

89 62 78 78

5 26.5 14 10.5

a The data were derived by integration of the two-dimensional

Percentage large cells

WM 5.5 11.5 8 11.5

15.5 42 24 22

histograms in Fig. 2.

RESULTS As noted previously (l-7), inclusion of gangliosides in the culture medium results in effective suppression of the proliferative responseof murine splenocytesto Con A (Fig. 1) without affecting cell viability. All gangliosides tested were suppressive and there was no apparent relationship between the sugar sequence of the gangliosides and their efficacy in suppressingproliferation. The amount of inhibition obtained in different experiments was somewhat variable but GDlband SPG were the most effective inhibitors, followed by GoLa,mixed bovine brain gangliosides,GTib, and GM,, in order of decreasingeffectiveness.Incorporation of gangliosidesinto liposomes that contained cholesterol and phosphatidylcholine or into a reconstituted lipoprotein particle that contained Apo A- 1,phosphatidylcholine, and gangliosidesincreasedthe efficiency of ganglioside inhibition (data not shown). E&et of gangliosides on the cell cycle.In order to examine parameters other than DNA synthesis,the effect of gangliosides on cell sizeand RNA content was evaluated by flow cytometry. After 24 hr in culture the effect of gangliosides on these parameters, and on the percentage of cells in different phasesof the cell cycle,was minimal.

‘=M

01 0

FlG.

25 50 76 100 GANGLIDSIDES f&ml)

I

125

1. Inhibition by gangliosides of concanavalin A-stimulated [3H]thymidine incorporation.

GANGLIOSIDE

INHIBITION

RNA

OF PROLIFERATION

75

RNA

DNA

RNA

DNA

RNA

DNA

DNA

FIG. 2. Flow cytometric analyses of the DNA and RNA content of spleen mononuclear cells after 48 hr of stimulation by Con A. The Xand Y axes represent integrated linear fluorescence. (A) Cells incubated in medium without stimulation; (B) Con A-activated lymphocytes; (C) Con A plus GM,, 200 &ml; (D) Con A plus Go,,, 200 &ml.

At 48 hr, however, more of the ganglioside-treated cells wre in Go/G,, these cells contained more RNA, and fewer cells were in S phase (Table 2, Fig. 2). In addition, fewer ganglioside-treated cells had enlarged. These data indicate that gangliosidetreated splenocytes were arrested in Go/G1 and few cells entered S phase. Expression of IL-2 receptors and binding of IL-2. The expression of IL-2 receptors on Con A-treated cells was evaluated by measuring the binding of a rat monoclonal antibody against the IL-2 receptor. After 24 hr in culture with Con A there was no difference in the number of IL-2 receptors on untreated and ganglioside-treated cells (Fig. 3). Inclusion of 10-1000 U/ml of IL2 in the culture medium had no effect on the ganglioside-induced inhibition of proliferation (data not shown). Two classes of IL-2 receptors have been described on activated human and murine T lymphocytes, high-affinity receptors that comprise less than 15% of total receptor sites, and more numerous low-affinity receptors (15). The physiological effects of IL-2 on activated cells are thought to be a consequence of the occupancy of highaffinity receptor sites by IL2 (15, 18-20). The concentration of IL-2 used in our studies, 100 PM, should represent binding only to high-affinity receptors (15). The binding of IL2 to 48-hr Con A blasts was effectively inhibited by Gnla (Fig. 4) but not by asialo GM, (data not shown). The inhibition was much more effective if the gangliosides were preincubated with IL-2 before addition of cells. No inhibition of

LOG FLUORESCENCE

3. Binding of anti-IL-2 receptor antibody 7D4 to (1) splenocytes incubated in medium, (2) incubated with Con A, (3) cells incubated with Go,,, 200 &ml, and (4) Con A plus Go,,. FIG.

76

MARCUS

OO

I loo

ET AL.

I 200 ODlr

1 300

I 400

(&ml)

FIG. 4. Inhibition of “‘1 K-2 binding to 48-hr activated Con A splenocytes. (O-O) Gangliosides and IL-2 were preincubated for 30 min at room temperature before addition to 48&r activated Con A blasts. (0 --- 0) Gangliosides and IL-2 were mixed and added to the activated cells immediately.

IL-2 binding was observed if cells were preincubated with gangliosides for 30 min and washed prior to addition of K-2. The binding of 12SI-labeled IL-2 could be inhibited completely by addition of unlabeled K-2. DISCUSSION The data presented above and previous reports ( 1, 12) indicate that certain early events in T-cell activation, increase in cell size, RNA synthesis, expression of the IL-2 receptor, and secretion of IL-2, are not substantially inhibited by gangliosides. The major effect of gangliosides is to prevent entry of the cells into S phase, apparently by interfering with the binding of IL-2 to cell membrane receptors. This hypothesis is supported by direct evidence for the interaction of IL-2 with gangliosides, namely the binding of IL-2 to immobilized gangliosides (11) and the ability of gangliosides to interfere with the binding of IL2 to anti-IL-2 antibodies (12). In addition, there is no evidence that the decreased binding of IL-2 to activated cells is caused by alteration of the cell membrane receptor by gangliosides. Cells that are incubated with gangliosides and then washed exhibit normal binding of both IL-2 and antibodies against the IL-2 receptor ( 12, and this report). This hypothesis concerning the mechanism of ganglioside inhibition is also consistent with the concept that the function of IL-2 is to mediate the transition of stimulated T cells from G, into the replicative phases of the cell cycle ( 19-2 1). Preincubation of IL-2 with gangliosides was required for effective inhibition of IL-2 binding to activated cells ( 12, and Fig. 4). The requirement for preincubation is probably a reflection of the affinity of IL-2 for high-affinity receptors (&of approximately 5-l 5 PM) and the low affinity of IL-2 for gangliosides, since a IO6 molar excess of gangliosides was required for half-maximal inhibition of IL-2 binding ( 12). Since the binding of gangliosides to cells is rapid, reaching a maximum within 30-60 min (3,22-24), it is curious that inhibition of mononuclear cell function can be demonstrated only following preincubation with gangliosides for periods of 48 hr

GANGLIOSIDE

INHIBITION

OF PROLIFERATION

77

or longer prior to addition of antigens or mitogens (3,7). These observations suggest that the inhibition observed during a 48-hr culture period, as in the experiments presented above, is caused by gangliosides in the medium and not by cell-associated gangliosides. Gangliosides have multiple effects on the metabolism and function of neuronal cells (25), including alterations of glycoconjugate metabolism and stimulation of axonal sprouting. The mechanisms by which gangliosides incorporated in the medium suppress cellular proliferation may differ from those in effect when mononuclear cells are preincubated with gangliosides for prolonged periods. If the binding of IL-2 is the sole or principal mechanism of ganglioside inhibition, the inhibition should be reversible by addition of an excess of IL-2. This was observed with murine IL-Zdependent cell lines and 5-day PHA blasts, but not with fresh PHA-stimulated mononuclear cells (12) or with Con A-stimulated mouse splenocytes in our experiments. These observations suggest that gangliosides may accomplish their inhibition by more than one mechanism, but experimental differences among these systems may account for the divergent results. The reversal of ganglioside inhibition was demonstrated in systems where gangliosides and IL-2 were mixed and preincubated prior to addition to activated cells, i.e., murine cell lines or PHA blasts, which were then placed again in culture. In the other experiments inclusion of IL-2 in the culture medium during the activation process did not overcome the ganglioside effect. IL-2 is readily inactivated by hydrolytic enzymes and the accumulation of these enzymes in tissue culture media, particularly during the prolonged culture periods required for human mononuclear cells, may result in the inactivation of the added IL-2. What are the physiological implications of the binding of IL-2 to gangliosides? It has been suggested that gangliosides, or glycoproteins that contain similar carbohydrate structures, are physiological receptors for IL-2, but this seems highly unlikely for a number of reasons. As discussed by Robb (12), the human lymphocyte IL-2 receptor (Tat) has been well characterized. It is a 55,000-kDa molecular mass integral membrane glycoprotein (26) whose gene has been cloned and expressed in murine L cells (27-30). Of particular importance to this discussion, IL-2 binds effectively to an unglycosylated precursor of Tat (31). Finally, the interaction of gangliosides with IL-2 is characterized by low affinity and a virtual absence of specificity for carbohydrate sequences. All gangliosides tested inhibited mononuclear cell proliferation regardless of the neutral sugar sequence present in the backbone structure, or the number or position of the sialic acid substituents. Since neither free sialic acid nor neutral glycosphingolipids are inhibitory; the only requirement for inhibition appears to be the presence of one sialic acid on a glycosphingolipid structure. In view of the presence of gangliosides on the surface of all cells, and the highly regulated and transient expression of functional IL-2 receptors, it is unlikely that gangliosides are physiological receptors for IL-2. A number of investigators have proposed that gangliosides may be immunomodulatory in vivo because of their ability to suppress mononuclear cell proliferation in vitro. As discussed in more detail elsewhere (8), the in vitro studies have employed gangliosides either as pure dispersions or in liposomes that contain a high mole percentage of gangliosides, and there is no evidence that gangliosides occur in this form in vivo. Gangliosides are known to occur only in cell membranes and in lipoproteins, where their molar concentration relative to other lipids is very low, and their accessibility to ligands such as antibodies may be limited. Statements about the immunosuppressive properties of gangliosides that are shed by tumor cells (6, 7) are misleading

78

MARCUS

ET AL.

becausethe shed material, presumably fragments of cell membranes, was not actually tested.The gangliosideswere extracted from this material, purified, and testedin the usual in vitro systems.Thus, there is no evidence that gangliosides in their natural environment, cell membranes and lipoproteins, are immunosuppressive in vitro or in vivo. On the contrary, neither patients nor experimental animals who have received daily injections of gangliosides for months exhibit any adverse systemiceffects (25). Further insight into the immunoregulatory properties of gangliosides will be gained only by experiments conducted in a biological setting. ACKNOWLEDGMENTS The authors thank Dr. James W. Thomas and Dr. Richard J. Robb for helpful discussions, and Ms. Charlene Shackelford for secretarial assistance in preparing the manuscript.

REFERENCES Lengle, E. E., Krishnaraj, R., and Kemp, R. G., CancerRes.39,871, 1979. Ryan, J. L., and Shinitzky, M., Eur. J Immunol.9,171, 1979. Whisler, R. L., and Yates, A. J., J. Immunol.125,2106,1980. Yates, A. J., Hitchcock, C. L., Stewart, S. S., and Whisler, R. L., In “Cell Surface Glycolipids” (C. C. Sweeley, Ed.), pp. 419-433. ACS Symposium Series, No. 128, American Chemical Society, 1980. 5. Krishnaraj, R., Lengle, E. E., and Kemp, R. G. Eur.J CancerClin.Oncol.l&89, 1982. 6. Ladisch, S., Gillard, B., Wong, C., and Ulsh, L., Cancer Res.43,3808,1983. 7. Lad&h, S., Ulsh, L., Gillard, B., and Won& C., J. Clin.Invest.74,2074,1984. 8. Marcus, D. M., Mol. Immunof.21, 1083,1984. 9. Krishnaraj, R., Lin, J., and Kemp, R. G., Cell Zmmunol. 78, 152, 1983. 10. Merritt, W. D., Bailey, J. M., and Pluznik, D. H., Cell Zmmunol. 89, 1, 1984. 11. Parker, J., Caldini, G., Krishnamurti, C., Ahrens, P. B., and Ankel, H., PEBSLett.170,391,1984. 12. Robb, R. J.,J. Immunol.136,971, 1985. 13. Kundu, S. K., Suzuki, A., Sabo, B., McCreary, J., Niver, E., Harman, R., and Marcus, D. M., J. Immunogenet. 8,357,1981. 14. Malek, T. R., Robb, R. J., and Shevach, E. M., Proc.Natl.Acad.Sci.USA80,5694,1983. 15. Robb, R. J., Greene, W. C., and Rusk, C. M., J. Exp.Med. 160,1126,1984. 16. Braylan, R. C., Benson, N. A., Nourse, V., and Kruth, H. S.,Cytometry2,337,1982. 17. Wallen, C. A., Higashikubo, R., and Dethlefsen, L. A., Cytometry3,155, 1982. 18. Dean, P. N., and Jett, J. H., J. CellBiol.60,523,1974. 19. Robb, R. J., Munck, A.,andSmith, K. A., J. Exp.Med.154,1455,1981. 20. Cantrell, D. A., and Smith, K. A., Science 224,1312, 1984. 2 1. Robb, R. J., Immunol.Today5,203, 1984. 22. Callies, R., Schwanmann, G., Radsak, K., Siegert, R., and Wiegandt, H., Eur. J. B&hem. 80,425, 1977. 23. Radsak, K., Schwarzmann, G., and Wiegandt, H., Hoppe-Seylers Z. Physiol.Chem.363,263,1982. 24. Schwarzmann, G., Hotfmann-Bleihauer, P., Schubert, J., Sandhoff, K., and Marsh, D., Biochemistry 1. 2. 3. 4.

22,5041,1983. 25.

26. 27. 28. 29.

Gorio, A., and Haber, B., Eds. Neurobiology of Gangliosides. Liss, New York, 1985. Robb, R. J., andGreene, W. C., J. Exp.Med.158,1332,1983. Leonard, W. J., Depper, J. M., Crabtree, G. R., Rudikoff, S., Pumphrey, J., Robb, R. J., Kronke, M., Svetlik, P. B., Peffer, N. J., Waldmann, T. A., and Greene, W. C.,Nature(London) 311,626, 1984. Nikaido, T., Shimizu, A., Ishida, N., Sabe, H., Teshigawara, K., Mae&, M., Uchiyama, T., Yodoi, J., and Honjo, T., Nature(London) 311,63 1, 1985. Greene, W. C., Robb, R. J., Svetlik, P. B., Rusk, C. M., Depper, J. M., and Leonard, W. J., J. Exp.

Med.162,363,1985. 30. Leonard, W. J., Depper, J. M., Kanehisa, M., Kronke, M., Peffer, N. J., Svetlik, P. B., Sullivan, M., and Greene, W. C., Science 2X),633,1985. 31. Leonard, W. J., Depper, J. M., Kronke, M., Robb, R. J., Waldmann, T. A., and Greene, W. C., J.

Biol.Chem.260,1872,1985.