Eur. J. Immunol. 1992. 22: 2783-2787

NBNTcells switch B cells to TgGJIgE synthesis

2783

Fritz Ledermann., Christoph Heusser., Claude SchliengerAand Graham Le Gros.

Interleukin-3-treated non-B, non-T cells switch activated B cells to IgGl/IgE synthesis

Research Department., Pharmaceutical Research, CIBA-GEIGY Ltd. and Basel Institute for ImmunologyA, Basel

The switch of activated B cells to IgE synthesis i s an interleukin (IL)-3-dependent process. It is currently thought that specificT cells activated by antigen presented in the context of class I1major histocompatibility complex are the major source of IL-4. Recently it has been demonstrated that a splenic non-T non-B cell population (termed NBNT) has the capacity to produce IL-4 following IgE and IgG receptor cross-linkage. In this study we demonstrate that IL-4 producing NBNT cells can induce the switch of lipopolysaccharide-activated B cells to the synthesis of IgGl and IgE antibodies. Furthermore, it was found that not only IgE receptor cross-linkage but IL-3 was able to stimulate NBNTcells to produce IL-4 and induce the switch of B cells to IgE synthesis. NBNTcells derived from the spleen and bone marrow of SCID mice were able to produce 1L-4 on exposure to IL-3. This suggested that the ability of IL-3 to stimulate IL-4 production was not dependent on prior exposure of the NBNTcells to antibody complexes in vivo. Taken together these findings represent the first observation that enough IL-4 is produced by NBNT cells to actually influence a B cell IgG/Ig response. The findings also clearly demonstrate that B cells do not need high concentrations of IL-4 to be directed to switch to IgGl and IgE synthesis.

1 Introduction

2 Materials and methods

IL-4 has been demonstrated to induce B cells to switch to IgE production [1-41. IL-4 production has recently been described in a non-T cell population present in spleen, bone marrow and lung tissue in response to IgE receptor cross-linkage [5, 61. This IL-4-producing cell population was characterized as expressing the high-affinity receptor for IgE and being devoid of T, B, M a and NK cell markers. Such cells were termed non-B non-T (NBNT) cells. In addition certain mast cell lines have been shown t o produce 1L-4 upon IgE receptor cross-linkage [7, 81.

2.1 Animals BALB/c, C57BL/6 and CB-17 SCID mice were obtained from the Ciba-Geigy Tierfarm (Sisseln, Switzerland). Infection of mice with Nippostrongylus brusiliensis by subcutaneous injection of 700 N . brusiliensis larvae. Mice were maintained in the Pharma Dept. animal facility and were used at 2 to 6 months of age.

2.2 Preparation of NBNT cells A further point indicated in the studies on NBNTcells was that the lymphokine IL-3 had a profound enhancing effect on the amount of IL-4 produced by IgE receptor crosslinked cells [9]. It was suggested that IL-3 could play a role in vitro and in vivo in regulating IL-4 production by its action on NBNT cells. In this report we demonstrate that NBNTcells isolated from the spleens of SCID [lo], normal or immunized mice can be activated by 1L-3 alone to induce the switch of LPSstimulated B cells to IgE synthesis. We used a sensitive and specific culture system containing the stromal cell line CB 5.1 as a feeder layer for IL-4-induced IgE production by B cells [11].The in vivo context in which B cell differentiation may be influenced by IL-3-activated NBNT cells is discussed.

Cell suspensions were prepared from spleens of untreated mice or mice that had been previously infected with N . brusiliensis. Red blood cells were lysed with Trisammonium chloride lysing buffer. Cells were washed twice in Hanks' balanced salt solution and suspended in culture medium.They were incubted at 4 "C for 1 h in the presence of the anti-FcylUI antibody 2.4G2 (10 pglml) [12] with a mixture of fluoresceinated antibodies containing 10 pg/ml of anti-Thy-1.2 (30H12) [13], anti-CD3 (2Cll) [14], antiCD4 (Gk1.5) [15], anti-CD8 (2.43) [16], anti-B220 (6B2) [17], and anti-Iad (MKD6) [HI. A t the end of the incubation, the cells were washed twice in staining buffer and mixed for 1h at 4 "C with a suspension of magnetic beads coated with sheep anti-fluorescein (FL) antibodies (Advanced Magnetics Inc., Cambridge, MA) (12 ml/lOs stained cells). Positively staining cells were depleted by two cycles of a 20-min exposure to a magnetic field. The remaining cells were washed twice in culture medium and examined for the removal of both B and Tcells by fluorescent microscopy.

[I 106381 Correspondence:Graham Le Gros, CIBA-GEIGY Ltd., Pharmaceuticals Research, R-1056.4.70, CH-4002 Basel, Switzerland

Abhreviation: NBNT Cells: Non-B. non-T cells

0 VCH Verlagsgesellschaft mbH. D-6940 Weinheim, 1992

2.3 B cell culture conditions C57BL/6 mouse spleens were used as a source of B cells. B cell cultures were established in flat-bottom wells of a 0014-2980/92/1111-2783$3.50+ .25/0

2784

Eur. J. Immunol. 1992. 22: 2783-2787

F. Lederrnann, C. Heusser, C. Schlienger and G. Le Gros

96-well plate (Nunc, Denmark) as described [11]. Each well contained medium (100 pl); 1000 CS7BL/6 spleen cells, 3500 CBS. 1fibroblast feeder cells [9] and 10 pg LPS ( E . coli 0.26 : B6, Difco Laboratories. Detroit, MI) with graded numbcrc of NBNTcells and various concentrations of either medium, IL-4, or IL-3 in a final volume of 200 pl. Cultures were incubated at 37 "C, 5% COz 97% humidity for 7 days. The culture supernatants were frozen until assayed.

HC1, pH 9.8 for 30 min.The absorbance was read at 405 nm in a Multiskan photometer (Flow, Irvine, Scotland). O n each plate an appropriate standard curve was included. For this purpose the following monoclonal mouse antibodies were used: 12-3 IgE (anti-PC); Klon 16 IgGl (anti-BPO); 23-1-1 IgM (anti-PC).

3 Results 2.4 Culture medium, growth factors and monoclonal antibodies

Dulbeco's modified Eagle medium supplemented with 10% preselected fCS, L-glutamine (2 mM), 2-mercaptoethanol (SO pM), gentamkin (100 pg/ml), sodium pyruvate (1 mM) and 1% of a cocktail consisting of folk acid (600 pg/ml), i>-asparagine (3.6 mg/ml) and L-arginine (11.6 mg/ml) was used for culturing cells. All ingredients were purchased from Gibco, Paisley, Scotland. Growth factors were prepared from transfected X63Ag8653 myelomas constitutively secreting large quantities of IL-3 or 1L-4 1191. IL-3 activity was standardized using the IL-3-dependent B13 cell line bioassay [20]. Here 0.3 unit of activity was defined as the concentration of lymphokine which produces half maximal proliferation of the B13 cell line. IL-4 activity was standardized using an IL-4 responsive CTLL cell line [21]. Three units of activity was defined as the concentration of lymphokine which produces halfmaximal proliferation of the CTLL cell line. The neutralizing rat anti-mouse IL-4 monoclonal antibody llB11 was purified from ascites fluid and added to cultures at a concentration of 10 pg/ml.

3.1 JL-3-treated NBNT cells from SCJD mice induce LPS-activated B cells to switch to IgGl and IgE synthesis NBNT cells were purified from the spleens and bone marrow of SCID mice and titrated into wells containing B cells, LPS and the CB5.1 stromal cell 1ayer.The cells were cultured for 7 days after which time supernatants were harvested and the concentrations of IgM, IgG and IgE isotypes determined.

-

In B cell cultures stimulated with LPS, 9 pg/ml IgM and 60 ng/ml IgGl but undetectable levels of IgE (< 1 ng/ml) were measuredin the supernatants (Fig. 1).The addition of NBNTcells from either the spleen or the bone marrow to the LPS-activated B cells had no effect on the immunoglobulin isotype pattern observed in the supernatants. However, when IL-3 was included in the culture medium, a

250

1'

I

Granulocyte (G)-CSF (R & D systems, Mineapolis, MN), granulocyte-macrophage (GM)-CSF, (macrophage) MCSF (Genzyme), IL-1 (Ciba-Geigy), IL-7 (kindly provided by R. Palacious, Basel Institute for Immunology), were prepared and used as indicated by the supplier. ['HIThymidine (1 FCi/well; specific activity S mCi/mmol) was present during the final 6 h of culture. Incorporated radiolabel from harvested cells was measured by liquid scintillation. The SD of triplicate cultures did not exceed 12% of the mean.

I

"a"

"

"

I '

"

"

"

"

' 1

2.5 Jsotype-specific ELISA The isotypes of the secreted mouse antibodies were determined by ELISA. Antibody-containing supernatants (SO pl) were tranrferred to 96-well plates (Immunoplate maxisorp F96, Nunc, Denmark) which had been coated (1 pglwell) with one of the following isotype-specific monoclonal rat antibodies: b-7-6 (anti-IgM, [22]); LOMGI-2 (anti-IgG1, Zymed, San Francisco, CA); 4B39 (anti-IgE, [ll]). Bound mouse antibodies from the culture supernatants were detected by biotinylated (a) b-7-6 (anti-IgM), (b) rabbit anti-mouse IgGl (Zymed, batch +k 70414), (c) 4A-1-28 (anti-IgE, [11]) and (d) alkaline phosphatase conjugated to avidin (Zymed).The amount of bound mouse immunoglobulin was revealed by incubating each well with 150 pl of 1 mg/ml p-nitrophenyl-phosphate (Sigma, St. Louis, MO) in 9.7% (v/v) diethanol amine-

0

3

10

60

300

SCID SC x 10 -2

0 3

10

60

SCID BMC x 10

300

-*

Figure 1. The effect of NBNT cells and IL-3 on LPS-stimulated B cell immunoglobulin synthesis. NBNT cells (0 to 3 X lo4 cells/well) prepared from SCID spleen cells (SC) or bone marrow cells (BMC) were titrated into cultures containing spleen B cells (1000/well), CB5.1 cells (3500/well) and LPS (10 pg/well). Cultures were supplemented either with nothing (0-O), 3000 U/ml IL-3 (W-R) or with 100 U/ml IL-4 (A).After 7 days of incubation, cultures were analyzed for secreted &GI, IgE and IgM antibodies.

Eur. J. Immunol. 1992. 22: 2753-2787

NBNTcell-dependent increase in IgGl(500 ng/ml) and IgE (60 ng/ml) synthesis was detected (Fig. 1). NBNT cells isolted from the bone marrow and activated by IL-3 stimulated peak IgE production at a lower cell concentration (104/well), although they became inhibitory at higher cell concentrations. The increase in IgGl, IgE isotypes was accompanied by a decrease (20-fold) in the level of IgM produced by the B cells.The addition of 1L-4to the cultures also stimulated the production of IgGl and IgE isotypes which occurred in both the presence and absence of NBNT cells. The IgE-inducing activity of 3-day culture supernatants harvested from cultures containing splenic NBNTcells (5 x 1Os//ml) from SCID mice was dependent on the presence of IL-3. Such supernatants (diluted 50%) induced the synthesis of 350 ng/ml IgE by LPS-stimulated B cells. Thus, it appeared that either a combination of NBNTcells and IL-3 or IL-4 alone could induce an identical immunoglobulin isotype pattern in LPS-activated B cells. This result indicated that the IL-3, NBNTcell effect might be IL-4 mediated. 3.2 Immunization enhances the capacity of NBNT cells to switch B cells to IgE synthesis

NBNT cells switch B cells to IgG,/IgE synthesis

2785

mice secreted enough IL-4 in the culture system to induce IgE secretion from 1000 LPS-activated B cells. Approximately 4-fold more NBNT cells were required for IgE induction when prepared from normal mice and 30-fold more cells were required when NBNT cells were prepared from SCID mice. This result indicates that the immune status of mice has a direct effect on the IL-4 producing capacity of the NBNT cells. Furthermore, if anti-IL-4 neutralizing antibody was added to replicate cultures, B cell differentiation to IgE secretion was completely inhibited. This result identified the major activity involved in the IL-3-stimulated NBNT cell-mediated B cell switch to IgE synthesis as IL-4. The possibility that the IL-3-regulated activation state of SCID NBNT cells differed from NBNT cells taken from immunized normal mice was tested in the experiment outlined in Fig. 3. NBNT cells were prepared from mice infected 10 days previously with N . brasiliensis worms or from SCID mice and their ability to support B cell switch to antibody-forming cells when cultured with different concentrations of IL-3 was analyzed (Fig. 3). Identical concentrations of IL-3 (> 100 U/ml) were required to induce both populations of NBNTcells to support B cell switch to IgE. Surprisingly, 10-fold less IL-3 was required for NBNT cells to assist B cells in switching to IgGl synthesis. In summary, NBNT cells which have never been influenced by immune responses are as sensitive to IL-3 stimulation as those NBNT cells which have been exposed to an immune response.

It has been shown that prior immunization of mice markedly increases the capacity of NBNT cells to produce IL-4 in response to IgE receptor cross-linkage [6]. This suggests that the cellular interactions of mediators released as a consequence of an activated immune system may influence NBNT cell activity. Therefore, we compared the 1L-4 producing capacity of IL-3-activated splenic NBNT cells isolated from N . brusiliensis-infected mice, untreated 3.3 Only IL-3 induces NBNT cells to support B cell switching to IgE synthesis mice or mice lacking bothTand B cells (SCID) (Fig. 2). As few as 1000 NBNT cells from N . hrusiliensis-immunized The ability of other hemopoietic active cytokines to influence NBNTcell release of IL-4 in the co-culutre system was tested (Table 1).The cytokines GM-CSF (250 U/ml), M-CSF (25 U/ml), G-CSF (250 U/ml), IL-1 (1ng/ml) and

I

.^

‘-&

40

U

-

E 3

150

-

100

-

50

-

C

.-U

Y)

0)

2

I

m P

2

f

+

200

” ?

20

0 0 0.30.6 1

3

6 10 30 60100 I

NBNT SC x 10 -3

Figure2. The effect of immunization on the capacity of NBNT cells to switch B cells to IgE synthesis. Splenic NBNTcells isolated untreated BALB/c mice either from SCID mice (H-m), (A-A) or BALB/c mice infected with N . brusiliensis (0-0) were addcd to cultures containing B cells. CB5.1 cells, LPS, and 1L-3 (300 U/ml). Replicate cultures also included the anti-IL-4 After 7 days culture antibody l l B l l (20 pg/well) (W---W). supcrnatants were analyzed for IgE by ELISA.

0

1

10

100

1000 100 0

IL3 (U/rnl)

Figure3. The effect of IL-3 concentration on the capacity of NBNTcells to switch B cells to IgE synthesis. IL-3 was titrated into B cell cultures supplemented with splenic NBNT cells isolated either from SCID mice (105/well(A)or BALBk mice infected with N . hrusiliensis (2.5 X 104/well) ( 0 ) .After 7 days of incubation, supernatants were analyzed for secreted IgGl (---), or IgE (-) by ELISA.

F. Ledermann. C. Heusser, C. Schlienger and G. Le Gros

2786

IL-7 (2500 U/ml) were added to co-cultures of NBNTcells, B cells (in presence of LPS) either seprately or together with IL-3 (300 U/ml). Only IL-3 was able to activate NBNT cells to support B cell switching to IgE synthesis. The possibility that the indicated cytokines could modify IL3-induced IgE production was also tested. GM-CSF, MCSF, G-CSF, IL-1 and IL-7 did not inhibit the IL-3-induced release of TL-4 by NBNTcells nor the switching of B cells to IgE by exogenous IL-4. 3.4 IL-4 production by NBNT cells can be induced by either 1L-3 or IgE receptor cross-linkage

In a previous study it was reported that cross-linkage of IgE receptors on NBNTcells leads to the production of IL-4. It was also reported that IL-4 release was enhanced in the presence of IL-3 but 1L-3 was unable to stimulate detectable IL-4 production by itself as analyzed by an IL4-dependent cell line. Therefore, it was important to compare 1L-4 production by either IgE- or IL-3-triggered NBNT cells measured by the IL-4-dependent B cell switch to IgE synthesis assay. NBNT from untreated mice cells triggered either with IL-3 (300 U/ml) or IgE bound to a plate secreted sufficient IL-4 to allow LPS-activated B cells to switch to IgE synthesis (Fig. 4A). The switch to IgE synthesis correlated with a significant inhibition of IgM formation (Fig. 4B). When the TL-4-dependent CTLL growth assay was used to measure IL-4 in these supernatants, only the NBNTcells on which the IgE receptor had been cross-linked produced significant levels of IL-4. Simultaneous treatment of 30 000 NBNT cells with IgE and IL-3 induced the release of 7-10 U/ml IL-4. NBNTcells triggered by IgE alone secreted 20%-30% less IL-4. Thus, NBNT cells cultured with IL-3

J

1

1

-

-

IL3

I

IgE lgE+IL3

Figure 4. The effect of IL-3 and IgE stimulation on NBNTcell IL-4 production measured by, switch to IgE production, and inhibition of IgM production. Three day culture supernatants were prepared form 30000 NBNT cells from untreated mice. These cells were stimulated either with nothing, IL-3, IgE complexes bound to a plate or a combination of IL-3 and plate bound IgE complexes. Culture supernatants were transferred (50% dilution) to a second culture containing B cell, CB5.1 feeder cells and LPS. After7 days supernatants were analyzed for the presence of IgE and IgM antibodics.

Eur. J. Immunol. 1992. 22; 2783-2787

Table 1. Only IL-3 induces NBNTcells to support B cell switching to IgE synthesisa)

Cytokine added GM-CSF M-CSF G-CSF IL1 IL-7 IL-3

IgE production (nglml) no IL-3 + IL-3 0 0 0 0 0 320 -

350 230 330 350 320 320 320

a) The indicated cytokines were added to LPS-activated B cell, NBNTcell cultures either in the presence or absence of IL-3. After 7 days culture the amount of secreted IgE was measured by ELISA.

produced insufficient IL-4 to be measured in the CTLL assay. By using our LPS-activated B cell culture system for detection of IL-4 activity [ll]it was also observed that IL-3 alone could induce IL-4 release of 2-4 U/ml IL-4, by NBNT cells.

4 Discussion This report establishes that NBNTcells exposed to IL-3 or IgE receptor cross-linkinginduce activated B cells to switch to IgE and IgGl synthesis, IL-3 was 4-fold better in triggering IL-4 production by NBNT cells than in supporting the growth of myeloid colony-forming precursor cells [23], data not shown). Furthermore, the culturing of NBNT cells with other defined hemopoietically active lymphokines did not result in IL-4 production, suggesting a unique role for IL-3 in stimulating NBNT cell production of IL-4. The degree to which NBNTcells supported B cell switching to IgE synthesis varied according to whether the NBNT cells were isolated from SCID mice, normal mice or mice infected with N . brusiliensis. It is not clear whether this finding reflects differences in the proportion of the IL4-producing cells present in the NBNT cell population or represent different stages of cell activation. The relatively low level of IL-4 produced by SCID NBNT cells might suggest that the level of circulating IgG and LgE immune complexes may modify the activation state of the NBNT cells, although an increase in numbers of NBNT cells synthesizing IL-4 in immunized mice would be at least as likely an explanation. The identification of the NBNTcell as a cell capable of inducing B cells to switch to IgE synthesis raises the important question of whether such cells function in vivo. There is no evidence for the participation of NBNT-like cells in the primary or secondary follicles of lymph nodes and spleen tissue of immunized mice. However, the gut mucosal tissues and bronchial-associated tissues of the lung are populated by mast cells and precursor mast cells which have properties similar to those of NBNTcells in the spleen and bone marrow [24]. Not only are the mast cells present in these tissues but they greatly expand in number and activity during the course of parasite-induced immune responses or

Eur. J. Immunol. 1992. 22: 2783-2787

responses involving polyclonal activation of IgE-producing B cells. One speculation would be that the NBNT cells migrate from the spleen and bone marrow to mucosal sites and form a regulatory part of the microenvironment of the mucosal tissues which influences IgE and IgA synthesis by a spectrum of lymphokines including IL-4, IL-5, IL-6 and IL-10 [4,25]. In this context, it is perhaps significant that in studies of parasitized mice, serum levels of IL-3 were demonstrated to determine the course of an immune response and consequently affect the outcome of the infection [26, 271. An important component of the in vivo IL-3-mediated effect may be to direct NBNT cells to produce cytokines in a manner similar to that reported in this in vitro study. In conclusion, this study demonstrates that both IL-3 and IgE receptor-mediated triggering of NBNT cells results in the synthesis of sufficient IL-4 to initiate IgE production by B cells. Although this observation appears to increase the complexity of interactions involving cytokines and the cells they act upon, it allows an appreciation of the mechanisms which might underly pathological processes such as allergic disease states. The initiation of an IgE response may be assisted by the action of Tcell-derived IL-3 on NBNTcells, while the subsequent maintenance and recruitment of further IgE-producing B cells may be positively regulated by IgE-mediated cross-linking of receptors on NBNTcells. Certainly, NBNTcells have the potential for a role, both in the amplification and prolongation of an IgE immune response. The authors thank Dr. B. Houlden for her helpful comments and

I! Meneghinello and R. Allenspach for their help in the preparation of thiy manuscript.

Receivcd May 12, 1992: in revised form June 26, 1992.

5 References 1 Coffman, R. L., Ohara, J., Bond, M.W., Carty, J., Zlotnik, A. and Paul. W. E., J. Immunol. 1986. 136: 4538. 2 Bergstedt-Lindyvist, S., Moon, H. B., Person, U., Moller, G., Heusscr. C. H. and Severinson, E., Eur. J. Irnmunol. 1988. 18: 1073.

NBNT cells switch B cells t o IgGJIgE synthesis

2787

3 Finkelmann, F. D., Katona, I. M., Urban, J. F. Jr., Snapper, C. M., Ohara, J. andPaul,W. E., Proc. Natl. Acad. Sci. U S A 1985. 83: 9675. 4 Pau1,W. E., Cell 1989. 57: 521. 5 Ben-Sasson, S. Z., Le Gros, G., Conrad, D. H., Finkelmann, F. D. and Paul, W. E., Proc. Natl. Acad. Sci. USA 1990. 87: 1421. 6 Conrad, D. H . , Ben-Sasson, S. Z., Le Gros, G., Finkelmann, F. D. and Paul, W. E., J. Exp. Med. 1990. 171: 1497. 7 Plaut, M., Pierce, J. H., Watson, C. J., Hanley-Hyde, J., Nordan, R. €? and Paul, W. E., Nature 1989. 339: 64. 8 Wodnar-Fillipowicz, A , , Heusser, C. H . and Moroni, C., Nature 1989. 339: 150. 9 Le Gros, G., Ben-Sasson, S. Z., Conrad, D. H., Clark-Lewis, I., Finkelmann, F. D., Plaut, M. and Pau1,W. E., J. Irnrnunol. 1990. 1458: 2500. 10 Bosma, G. C., Custer, R . P. and bosma, M. J., Nature 1983. 301: 527. 11 Ledermann, F., Schlienger, F. ,Wagner, K. and Heusser, C. H.. J. Irnrnunol. Methods 1991. 141: 263. 12 Unkeless, J., Scgliano, E. and Freedman, V , Annu. Rev. Irnrnunol. 1988. 6: 251. 13 Ledbetter, J. A. and Herzenberg, L. A., Irnrnunol. Rev. 1979. 47: 63. 14 Leo, O., Foo, M., Sachs, D., Samelson, L. E. and Bluestone, J. A , , Proc. Natl. Acad. Sci. USA 1987. 84: 1374. 15 Dyalynas, D. I?, Quan, Z . S., Wall, K. A , , Pierres, A., Quintans, Loken, M. R., Pierres, M. and Fitch, F. W., J. Irnrnunol. 1983. 131: 2445. 16 Sarmiento, M., Glasebrook, A . L. and Fitch, F.W., J. Irnmunol. 1980. 125: 2665. 17 Coffman, R. L., Irnmunol. Rev. 1982. 69: 5. 18 Kappler, J.W., Skidmore, B. ,White, J. and Marrack, P., J. Exp. Med. 1981. 153: 1198. 19 Karasuyama, H . and Melchers, F., Eur. J. Irnrnunol. 1988. 18: 97. 20 Rolink, A. G., Melchers, F. and Palacios, R., J. Exp. Med. 1989. 169: 1693. 21 Hu-Li, J., Ohara, J., Watson, C., Tsang, W. and Paul, W. E., J. Immunol. 1989. 142: 800. 22 Julius, M. J., Heusser, C. H. and Hartmann, K . U., Eur. J. Irnrnunol. 1984. 14: 753. 23 Stephenson, J. R., Axelrad, A. A., McLeod, D. L. and Shreeve, M. M., Proc. Natl. Acad. Sci. USA 1971. 68: 1542. 24 Guy-Grand, D., Dy, M., Luffau, G . and Vassalli, P., J. Exp. Med. 1984. 160: 12. 25 Moore, K. W.,Viera, P., Fiorentino, D. F.,Trounstine, M. L., Khan, T. and Mosmann, T. R., Science 1990. 248: 1230. 26 Feng, Z.Y., Louis, J., Kindler,V , Pedrazzini, T., Eliason, J. F., Behin, R . and Vassalli, l?, Eur. J. Irnrnunol. 1988. 18: 1245. 27 Filho, M. A , , Dy, M., Lebel, B., Luffau, G. and Hamburger, J., Eur. J. Irnmunol. 1983. 13: 841.