Role of IL5 and IL2 in antigen-specific antibody responses

Eur. J. Immunol. 1990. 20: 2219-2227

Rosemarie H. DeKruyff, Rim R. Mosmann+ and Dale T. Umetsuov Department of Pediatricso, Stanford University School of Medicine, the Children's Hospital at Stanford, Stanford and DNAX Research Institute for Molecular and Cellular Biology+, Palo Alto

2219

Induction of antibody synthesis by CD4+ T cells: IL 5 is essential for induction of antigen-specific ~ not T H clones* ~ antibody responses by T H but Murine B cells stimulated with lipopolysaccharide (LPS) and interleukin (IL) 4 produce IgGl and IgE, but synthesize IgG2, when stimulated with LPS and interferon-y. The cytokines, however, that regulate immunoglobulin (Ig) synthesis induced in normal B cells under antigen-driven major histocompatibility complex (h4HC)-restricted conditions in the absence of potent B cell mitogens have not been fully elucidated. We and others have shown that under cognate MHC-restricted conditions, CD4+ Tcell clones of the T H 1 subset, which produce IL2 and interferon-y, and Tcell clones of the T H subset, ~ which produce IL4 and IL 5, are both capable of inducing anti-trinitrophenyl IgG plaque-forming cells. In this report we have examined in further detail the cytokine requirements for the induction of Ig synthesis in B cells cultured directly with T H and ~ T"2 Tcell clones. Using (a) T Hclones ~ that varied in the amount of IL5 secreted, (b) a neutralizing monoclonal antibody against IL 5 and (c) Tcell clones pretreated with cyclosporin A to inhibit cytokine secretion, we found that IL5 was essential for induction of IgGl synthesis byTH2 but not T HTcells. ~ Although we demonstrated that IL2 could actually up-regulate the synthesis of IL5 by T Hclones, ~ the induction of IgG synthesis by T H clones ~ was entirely independent of IL2. In contrast, induction of IgGl synthesis by T H clones ~ was absolutely dependent upon the presence of IL2 and was not affected by the presence of ILS.Thus, these studies demonstrate the idea that at least two independent pathways exist for the induction of IgGl synthesis, and that one of these pathways is IL4/IL5 dependent and the other IL2 dependent.

1 Introduction T cell help is required for the induction of antibody synthesis. However, the precise cytokine requirements for the induction of isotype-specific Ig synthesis have not been fully delineated. B cells stimulated with LPS in the presence of IL4 have been shown to synthesize IgGl and IgE [l,21, while in the presence of IFN-y, such B cells have been shown to synthesize IgGza but not IgGl or IgE [3]. B cells stimulated with LPS and transforming growth factor-p produce IgA, but not IgGl or IgG2, [4]. Although IL2 and IL5 are important B cell growth and differentiation factors [5-151, both IL2 and IL5 have limited effects on B cells activated with LPS [4, 161.

cate that the cytokine requirements for Ig synthesis are more complex. We and others [17-19] have demonstrated that under cognate, h4HC-restricted conditions, CD4+ T cell clones designated as T H (which ~ produce IL2 and IFN-y and are responsible for DTH reactions [20]), as well as clones designated as T H (which ~ produce IL 4 and IL 5 , but not IL2 or IFN-y [21]), are both capable of inducing IgG anti-TNP PFC responses. Because the cytokine profiles of T Hand ~ T Hclones ~ are distinct, these results suggest that two independent pathways for the induction of IgG synthesis exist. However, the precise cytokines involved in each pathway are not known.

The purpose of our current studies was to examine in detail the cytokine requirements for induction of Ig synthesis in ~ T H clones ~ under Although studies with LPS-activated B cells have provided murine B cells cultured with T H and pertinent insights into the regulation of Ig synthesis by cognate MHC-restricted conditions, and to analyze the cytokines, studies of normal B cells stimulated by T cells precise role of IL2 and of IL5 in B cell differentiation. under antigen-driven MHC-restricted circumstances, indi- Using several different experimental approaches to isolate the effects of IL2 and of IL5, we demonstrated at B cells, which have undergone a cognate interaction with T cells, [I 85781 require the presence of both IL5 and IL4 in order to synthesize IgGl. However, we showed that via an IL4/IL5* This work was supported by National Institutes of Health grants independent pathway, IL2 can also induce IgGl synthesis in A1 24571, A1 26322 and R R 05353 from the U.S. Public Health such B cells. Although we clearly demonstrated that the Services, and a grant from the American Lung Association. presence of IL 2 can increase the synthesis of IL 5 in some Recipient of the Hyde Watson Award from the Asthma and TH2Tcell clones, the effect of IL2 on IgGl synthesis could Allergy Foundation. occur in the complete absence of IL 4 and IL 5.These results indicate that both IL2 and IL5 are potent B cell differenCorrespondence: Rosemarie H. DeKruyff, Department of Pediatrics, Room S 217, Stanford University Medical Center, Stanford, tiation factors, that both can independently regulate IgGl synthesis, and that IgGl synthesis can occur via a pathway CA 94305. USA involving IL4 and IL5, as well as a pathway involving Abbreviation: CsA: Cyclosporin A IL 2. ~

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990

0014-2980/90/1010-2219$3.50+.25/0

2220

R. H. DeKruyff,T. R. Mosmann and D.T. Umetsu

Eur. J. Immunol. 1990. 20: 2219-2227

2 Materials and methods

2.4 Medium and B cell culture

2.1 Animals and antigens

Cells were cultured in DMEM (Gibco, Grand Island, NY) supplemented as previously described [28,29], and containing 10% FCS (Hyclone Laboratories, Logan, UT). Splenic B cells were prepared by sequential positive and negative selection. B cells were positively selected by adherence to goat anti-mouse Ig-coated plates essentially as described [30]. After elution, cells were treated with anti-Thy-1.2 plus rabbit C (Pel-Freeze, Brown Dear,WI) to remove residual Tcells. B cells (3.5 x lo6) were incubated in DMEM (1 ml in 24-well plates) at 37 "C in 10% COz.Where indicated, splenic adherent cells (SAC; 3300 rad irradiated) were added to cultures as additional accessory cells. SAC were prepared by incubating 3.5 x lo6 irradiated spleen cells in complete medium per well. After 2 h, wells were washed with BSS to remove nonadherent cells (>98% of the cell population).

Mice were obtained from the Jackson Laboratory, Bar Harbor, ME, or from the Institute for Medical Research, San Jose, CA. P o l y ( ~ - G l u ~ ~ - ~ - L y s ~ ~(GL@; - ~ - P hMilesel~) Yeda, Rehovot, Israel), KLH (Calbiochem, San Diego, CA) and conalbumin (Sigma Chemical Co., St. Louis, MO) were conjugated with 2,4,6-trinitrobenzenesulfonicacid according to the method of Eisen et al. [22].

2.2 Immunizations and mAb Mice (B cell donors) were primed with 300 pg of TNPconalbumin in CFA 3 to 5 weeks prior to use. Anti-Thy-1.2 mAb [HO-13-4, No. TIB99, American Type Culture Collection (ATCC), Rockville, MD] was prepared from serum-free culture SN by ammonium sulfate precipitation. Purified rat anti-IL 4 mAb [23] was similarly prepared from l l B l l hybridoma cells, generously provided by Dr. J. Ohara and Dr. W. Paul (National Institutes of Health, Bethesda, MD). Anti-IL2R mAb 3C7 and 7D4 [24] were provided by Dr. E . Shevach, NIH. Anti-IL2R mAb PC61 5.3 [25] was obtained from ATCC (No.TIB 222). Anti-IL5 mAb were prepared as described [26]. A rat mAb specific for mouse IgE, EM95, generated b y Z. Eshhar [27] was provided to us by R. Coffman (DNAX Research Institute, Palo Alto, CA) and was purified from hybridoma SN by ammonium sulfate precipitation followed by chromatography on DEAE-Sephacel (Pharmacia, Piscataway, NJ). The mouse IgE-producing hybridoma IGELa2 was obtained from ATCC (TIB 169). Ascites fluid from this hybridoma was partially purified by ammonium sulfate precipitation, and quantitated against purified IgE antibody generously provided by R. Coffman. For measurement of IgG subclasses and IgA concentrations by ELISA, goat antibodies to IgGl, IgGZ,, IgGZb, IgG3 and IgA were purchased from Southern Biotechnology Associates Inc., Birmingham, AL.

2.3 Derivation of clones The derivation and maintenance of clones E l 0 [28], D3, E6, BRD2, AK1 and B5 [19] have been described. Clones DClO and DC12 were obtained from a DBA/2 mouse that had been immunized i.p. with 100 pg of KLH in CFA. After 4 weeks, the mouse was boosted with 100 pg KLH in PBS, and after 7 days the spleen was removed and a single-cell suspension was prepared. T cells were prepared as the nonadherent cells on goat anti-mouse Ig-coated plates, followed by treatment with anti-Ly-2.1 and C. DC12 was derived from this population of cells cloned directly in the presence of conditioned medium containing IL 2. DC12 arose from an initial cell density of 250 cells/well and was subsequently subcloned at 0.3 cells/well. Some of the KLH-primed Tcells were originally cultured in the presence of irradiated (2500 rad) syngeneic spleen cells and 50 pg/ml of KLH. Cells were later cloned in the presence of conditioned medium containing IL 2, as previously described [29]. DClO arose from such a cloning.

2.5 PFC assay Cultures were assayed on day 5 for PFC onT"-conjugated sheep erythrocytes as previously described [31]. Indirect PFC were developed by the addition of affinity-purified rabbit anti-mouse Ig (Jackson Immunoresearch, West Haven, PA). IgG PFC were calculated as total PFC (facilitated by polyspecific rabbit anti-mouse Ig) minus direct (IgM) PFC.TheTNP specificity of the PFC generated under these conditions was demonstrated by measuring PFC with unconjugated sheep erythrocytes and finding no measurable response. 2.6 Measurement of Ig isotype concentrations in SN

Culture SN to be assayed for Ig isotype concentrations were harvested 7 days after the initiation of the cultures and frozen until assayed. Ig isotype concentrations were measured by an ELISA as described [19]. For the IgE ELISA, plates were coated with anti-IgE mAb, EM95. After addition of IgE standards or of samples, biotin-conjugated EM95 was added as the second-step antibody. Horseradish peroxidase-conjugated strepavidin was then added before development with o-phenylene diamine. For IgA and IgG subclass assays, goat anti-IgA and anti-IgG subclassspecific antibodies were purchased from Southern Biotechnology Associates. 2.7 Lymphokines

rIL 5 in the form of a COS cell SN (containing 17.95 ng/ml IL5) was a generous gift from the Genetics Institute, Cambridge, MA, and used for most of the experiments. rIL 5 purchased from Genzyme (Boston, MA) was used for a few experiments. Human rIL 2 was obtained from Amgen (Thousand Oaks, CA). Murine rIL4 was obtained from Genzyme . 2.8 Lymphokine assays

LL2 and IL4 were assayed by using the T cell growth factor-dependent line HT2 (generously provided by Dr.

Eur. J. Immunol. 1990. 20: 2219-2227

Role of IL5 and IL2 in antigen-specific antibody responses

Sam Strober, Stanford). HT2 cells (1 x lo4) were added to dilutions of test samples in 96-well plates. Differential blocking of IL4 or IL2 was achieved by using the anti-IL4 mAb 11B11,or a cocktail of anti-IL2R mAb, 3C7,7D4 and PC615.3. After 18 h, cells were pulsed with 1 pCi = 37 kBq of [3H]dThd for 6 h. Cultures were harvested with a PHD harvester (Cambridge Technology, Cambridge, MA) and [3H]dThd was measured using standard liquid scintillation Table 1. Lymphokine synthesis of clones used in this studya) Clone

Stimulus

Activity m - Y

IL2

IL5

TH1clones

D3 D3 El0 El0

283

Con A Antigen Con A Antigen

8 215 13 IL4

TH2clones BRD2 DClO DC12 DC13 AK1

B5

4215 1335 2490 2635

0.05 0.09 0.20 0.01

m Y

IL5

14 750 4687 6900 49500 14400 6800

Antigen Antigen Antigen Antigen Antigen Antigen

135 ND ND ND 2.32 1.2

12.75 7.79 4.42 43.04 1.63 1.26

2221

counting techniques. Units of IL 2 present in clone SN were calculated using human rIL2 obtained from Arngen as a standard. Units of IL4 contained in clone SN were calculated using murine rIL4 obtained from Genzyme as a standard. However, a later batch of rIL4 obtained from this company had approximately 100-fold more activity per unit in its ability t o stimulate HT2 proliferation as well as its ability to induce IgE synthesis by LPS-stimulated B cells. Therefore, quantitations of IL4in clone SN tested using the latter batch as standard were normalized to the former batch to allow direct comparisons. IL5 and IFN-y were measured by ELISA as described [26, 321. 2.9 Tcell SN

SN were prepared either by antigen stimulation or Con A stimulation of cloned Tcells. Antigen-stimulated Tcell SN were generated by stimulation of 5 X lo5 cloned Tcells/ml for 24 h with 50-100 pg/ml of antigen and 3.5 x lo6 syngeneic spleen cells (irradiated 2500 rad). Con A-stimulated Tcell SN were prepared by stimulation of lo6Tcells/ml with 1 pg/ml Con A for 18 h. Residual Con A in SN was blocked by addition of 10 mg/ml aMM. For evaluation of the effect of rIL2 on IL5 production, clone B5 was stimulated withT cell-depleted spleen cells (irradiated 2500 rad) in the presence or absence of 10 U/ml rIL2. 2.10 CyclospoM A (CsA) treatment

a) Cells of the indicated clone were stimulated with irradiated syngeneic spleen cells and antigen (for clone E10, GLQ, 50 pg/ml; for all other clones, KLH, 100 @ml) or with Con A (2 pg/ml) where indicated. SN were tested for lymphokine production as described in Sect. 2.8. IL2 and IL4 are expressed as U/ml, IL5 and IFN-y in nglml.

Cloned T cells were treated with CsA essentially as described by Krusemeier et al. [33]. The CsA was a generous gift of Dr. David Winter, Sandoz Research Institute. CsAwas dissolved as described [33], and added at 10 pg/ml to cloned Tcells (5 x lo5).Tcells were incubated

Table 2. IL5 restores the ability of the non-helper clone B5 to induce antibody secretion by primed B cellsa) TI, cells added

Exp. 1 B5 B5 B5

-

Exp. 2 B5 B5

-

Exp. 3 B5 B5 B5 Exp. 4 AKl AKl

-

Added to culture Antigen Lymphokine

TNP-KLH TNP-KLH TNP-KLH

-

TNP-KLH TNP-KLH

-

TNP-KLH KLH TNp-Co+KLH TNP-KLH

TNP-KLH TNP-KLH

Exp. 5 AK1

TNP-KLH

AK1

TNP-KLH

-

BRD2SN rIL5 BRD2 SN rIL5

-

Anti-TNP PFclcUlture IgM IgG

260 4080 2980

0 7120

80

40 0

140

im

6940

rIL5 rIL5

3880

0 3240

40

0

rIL 5 rIL5 rIL5

4060 160 100

2100

rIL5 rIL5

-

3800 4680 120

-

4640

Anti-ILS

580

0 80 0

3290 0

ND ND

a) TNP-primed BALBlc (3.5 X lo6) cells were cocultured with 3 x 104clone B5 or clone AK1 Tcells in the presence of the indicated antigen [KLH, TNPKLH or TNP-conalbumin (TNP-Co), 10 nglml]. Where indicated, rIL5 (1 nglml), or BRD2 SN (containing > 4 nglml IL5) was added to the cultures. B cells plus antigen in the absence of T cells produced 0 to 20 PFC/culture; addition of rIL5 produced 80-100 PFC/culture. Anti-IL 5 mAb TRFK-5 (10 pg/ml) was added to the cultures where indicated.

2222

R. H. DeKruyff,T. R. Mosmann and D.T. Umetsu

with CsA for 24 h at 37"C, then washed and added to cultures of purified B cells in the presence of 20 ng/ml CsA.

3 Results 3.1 General remarks The experiments in this report were performed using a panel of murineTHl and T H CD4+ ~ clones which are listed in Table 1 along with their lymphokine profiles. Both T H clones E l 0 and D3 produced large quantities of I L 2 and negligible quantities of IL 5, when stimulated optimally with antigen or Con A.While all of t h e T ~ clones 2 produced significant quantities of IL4, there was a great variability in the amount of IL5 secreted by these clones: theTH2clones B5 and AK1 produced very little IL5, compared with the T H clones ~ BRD2 and DC10, and compared with other typical T H clones ~ (Mosmann, unpublished observations). Our examination of the role of IL5 in the induction of Ig synthesis was greatly facilitated by the availability of these T H clones ~ which varied in their capacity to secrete IL 5.

3.2 IL5 is required for the induction of IgG by T H ~ clones We first examined the induction of Ig synthesis using the T H clones ~ B5 and AK1, both of which produced IL4 but only small amounts of IL 5. Clone B5 failed to induce IgM

3000

f 2000

t a

1000

-a 9 Yn. o L

0

1:400

1:lOO

1:200

150

1:25

Eur. J. Immunol. 1990. 20: 2219-2227 or IgG anti-TNP PFC when cultured with purified syngeneic TNP-primed B cells and TNP-KLH (Table 2, lines 1 and 6). Table 2 shows that addition of rIL5 or SN from theTH2 clone BRD2 restored the helper function of clone B5 for the induction of both IgM and IgG PFC responses. rIL5 alone, in the absence of the Tcell clone, was incapable of stimulating antibody synthesis (Table 2, lines 5 and 8), indicating the absence of preactivated B cells in the starting B cell population [7]. Induction of Ig by clone B5 in the presence of IL5 was normal in all other respects and required cognate T-B cell interaction. Thus the reconstitu~tion of helper function of this KLH-specific clone with rIL 5 required carrier-hapten linkage with TNP-KLH (Table 2, Exp. 3). No response was observed in the presence of KLH plus TNP-conalbumin. The T H clone ~ AK1, which also secreted only modest amounts of IL5, was able to induce IgM but not IgG anti-TNP PFC responses when cultured withTNP-primed B cells (Table 2, Exp. 4). However, induction of IgM by this clone was dependent on IL 5, since addition of a neutralizing anti-IL 5 mAb (TRFK-5) inhibited the IgM response induced by this clone (Table 2, exp. 5). Addition of rIL 5 to cultures of clone AK1 and B cells resulted in induction of IgG as well as IgM anti-TNP PFC (Exp. 4). Although the amounts of IL5 produced by clone AK1 and B5 when stimulated maximally were very similar (Table l ) , clone B5 may produce less I L 5 than clone AK1 at the very low antigen concentrations used in the B cell cultures, or alternatively clone AK1 may produce more of other factors that synergize in the induction of IgM. Fig. l a shows that the number of anti-TNP PFC induced by clone B5 increased when more IL5 was present in the cultures. Low concentrations of IL5 resulted only in induction of IgM anti-TNP PFC, while higher concentrations of IL5 resulted in induction of both IgM and IgG responses. Similar results were seen with clone AK1 (Fig. lb), in that the number of IgG anti-TNP PFC observed increased as higher concentrations of IL 5 were present in the cultures. These data indicate that the failure of clone B5 and AK1 to help in IgG PFC responses was due to insufficient production of IL5, and that IL5 is essential for the induction of IgG and IgM PFC responses by T H ~ clones under MHC-restricted conditions.

3.3 Anti-IL 5 mAb inhibits induction of antibody synthesis by T Hbut ~ not TH1clones Since it has been suggested that small amounts of IL5 produced by T H clones ~ might be important in the induction of Ig by THl clones [34],we investigated the role of IL5 in the induction of IgG PFC by T H clones ~ using the anti-IL 5 mAb TRFK-5. Table 3 shows that addition of "n .' anti-IL 5 mAb to cultures of the T H clone ~ D3 and B cells 0 1:400 1:200 1:lOO 150 1:25 had no effect on the anti-TNP PFC response. Since the dilution of rlL-5 amount of anti-IL5 mAb added to these cultures was sufficient to neutralize I L 5 produced by the T H clone ~ Figure 1. Effect of addition of rIL5 on induction of antibody DC12 (Table 3, line 4) and since identical results were responses mediated by clones B5 (a) and AK1 (b). TNP-primed B cells (3.5 x 1s)were co-cultured with the indicated cloned Tcells obtained with clone D3 using larger amounts of anti-IL 5 in the presence of 10 nglml TNP-KLH.The indicated dilution of a mAb, it is unlikely that the small amounts of IL 5 produced COS cell SN containing 17.95 nglml IL5 was added to the cultures. by clone D 3 exceeded the neutralizing capacity of the IgM (El) and IgG (+)anti-TNP PFC are presented. Similar results anti-IL5 mAb in these cultures. Furthermore, addition of were obtained in two other experiments. IL 5 to cultures of T H clones ~ and B cells did not change the I

.

I

.

I

.

I

.

I

.

I

Role of IL5 and IL2 in antigen-specific antibody responses

Eur. J. Immunol. 1990. 20: 2219-2227

Table 3. Anti-IL5 antibody inhibits B cell responses induced by TH2but not by T H clonesa) ~ Th C e h added

D3 (TH1) D3 DC12 (T"2) DC12

Addition to culture

Anti-TNP PFuculture IQM

hG

-

5520

Anti-ILS

5500

13 120 13380

-

am

5440

Anti-IL 5

1160

1360

2223

addition of rIL2 or of rIL5 to cultures of clone I35 or AK1 and B cells in the presence of TNP-KLH greatly enhanced the IgG PFC response. These results indicated that IL2 couldsubstitute f& IL5, or alternatively that IL2 induced the secretion of IL5 in these clones.

a) TNP-primed BALBlc (3.5 X 106) cells were co-cultured with cells of the indicated clone (3 x 104 DC12; 0.5 X 104 D3). Cultures were stimulated withTNP-KLH (10 nglml with DC12 and 1 nglml with D3). Anti-IL5 mAb TEWK-5 (5 pg/ml) was added to the cultures where indicated.

distribution of IgG subclasses (IgGI remained the major subclass induced, Table 4), though levels of IgA antibody were somewhat increased in the presence of rIL5. These results taken together indicate that the induction of IgGl by T"1 clones occurs independently of IL 5. 3.4 Role of IL2 in the induction of Ig

We next wished to examine the role of other cytokines in the induction of Ig responses by T H clones. ~ Of the cytokines produced by T H clones, ~ IFN-y has been shown to enhance IgG2, synthesis, but is inhibitory for IgE and IgG3 [3,35]. On the other hand, IL2 is known to be an important B cell growth factor, and we asked if IL2 alone could provide the lymphokine signals necessary for the induction of Ig synthesis. Our approach in evaluating the role of I L 2 in Ig synthesis was to examine Ig synthesis induced by clones that did not secrete I L 2 (e.g.,T H clones ~ that were unable by themselves to induce Ig synthesis) and attempt to reconstitute the response with exogenous IL2. Fig. 2 shows that

additions to culture

Figure 2. IL2 or IL 5 restores the ability of clones B5 (a) and AK1 (b) to induce antibody secretion ( , IgM; W , IgG) by primed B cells. TNP-primed BALBlc (3.5 X lo6) cells were co-cultured with 3 x 104cloneTcells in the presence of TNP-KLH (10 ng/ml).Where indicated, rIL5 (1 nglml) or rIL2 (10 U/ml) was added to the culture. B cells plus antigen in the absence of Tcells produced 0 to 20 PFC/culture; addition of rIL2 did not increase the number of PFC; addition of rIL5 produced 80-100 PFUculture.

Table 4. Effect of addition of rIL5 on induction of antibody production by T H clonesa) ~

Exp. 1 El0 ( T H ~ ) El0

rIL5

1260 3560

-

-

rIL 5

60

m

Exp. 2 D3 ( T H ~ ) D3

rIL 5

-

1140 1700

rIL 5

60

17 860 14300 0 0

-

-

-

0

0

3 450 2 410 0

1518 1194 (20

909

592

660

493 < 10 < 10

1920 1089