JOURNAL OF INTERFERON AND CYTOKINE RESEARCH 21:971–980 (2001)Mary Ann Liebert, Inc.

Transforming Growth Factor-b Has Contrasting Effects in thePresence or Absence of Exogenous Interleukin-12 on the In Vitro Activation of Cells That Transfer Experimental

Autoimmune Thyroiditis

WESLEY FONDAL, JR.,1 CHRISTOPHER SAMPSON,1 GORDON C. SHARP,2,3

and HELEN BRALEY-MULLEN1,2,4

ABSTRACT

Mouse thyroglobulin (MuTg)-sensitized spleen cells activated in vitro with MuTg induce experimental au-toimmune thyroiditis (EAT) in recipient mice with a thyroid infiltrate consisting primarily of lymphocytes. Amore severe and histologically distinct granulomatous form of EAT (G-EAT) is induced when donor cells areactivated with MuTg together with anti-interferon-g (IFN-g), anti-interleukin-2 receptor (IL-2R) monoclonalantibody (mAb), and IL-12. Transforming growth factor-b (TGF-b) is a multifunctional cytokine that canboth suppress and exacerbate autoimmune diseases and often has inhibitory effects on lymphocytes. To de-termine if TGF-b could modulate the in vitro activation of effector cells for G-EAT, TGF-b was added to cul-tures of MuTg-sensitized donor spleen cells together with MuTg. Cells activated in the presence of 2 ng/mlTGF-b induced moderately severe G-EAT in recipient mice. G-EAT induced by cells activated in the pres-ence of TGF-b was histologically similar but less severe than the G-EAT induced by cells activated in thepresence of IL-12. IL-12 and TGF-b modulate the activation of G-EAT effector cells by distinct mechanisms,as cells activated by TGF-b could induce G-EAT in the presence of anti-IL-12, and TGF-b inhibited the ef-fects of IL-12 on EAT effector cells. TGF-b exerted its activity during the first 24 h of the 72-h culture, whereasIL-12 functioned primarily during the final 24 h of culture. These results indicate that thyroid lesions withgranulomatous histopathology can be induced by both IL-12-dependent and IL-12-independent mechanisms,and TGF-b can exert both positive and negative effects on the effector cells for G-EAT.

971

INTRODUCTION

EXPERIMENTAL AUTOIMMUNE THYROIDITIS (EAT) is a chronicinflammatory autoimmune disease characterized by infil-

tration of the thyroid by lymphocytes and mononuclear cells,resulting in destruction of thyroid follicles.(1,2) EAT can be in-duced in susceptible strains of mice by immunization withmouse thyroglobulin (MuTg) and adjuvant(1) or by transfer ofin vitro activated spleen cells from MuTg-immunized donormice into syngeneic recipients.(2) Donor spleen cells activatedwith MuTg alone induce a chronic lymphocytic form of EATcharacterized by infiltration of the thyroid by mononuclear cellsand production of anti-MuTg autoantibodies.(1,2) Spleen cellsactivated with MuTg together with interleukin-12 (IL-12),(3)

anti-IL-2 receptor (IL-2R),(4) or anti-interferon-g (IFN-g)(5)

monoclonal antibodies (mAb) develop a more severe and acutegranulomatous form of EAT (G-EAT). The thyroid lesions inG-EAT are characterized by follicular cell proliferation, withlarge numbers of histiocytes, multinucleated giant cells, andvariable numbers of neutrophils in addition to T lympho-cytes.(3,4)

Transforming growth factor-b (TGF-b) plays an importantrole in regulating immune responses and various infectiousand autoimmune diseases.(6,7) TGF-b can have both positiveand negative effects on cells depending on their state of dif-ferentiation and also on the cytokine milieu.(6) In vivo ad-ministration of TGF-b suppresses some autoimmune diseases,such as experimental autoimmune encephalomyelitis (EAE),collagen-induced arthritis (CIA), and diabetes in NODmice,(7–12) and endogenous TGF-b produced by regulatory T

Departments of 1Molecular Microbiology and Immunology, 2Internal Medicine, and 3Pathology, University of Missouri School of Medicine,Columbia, MO 65212, and 4VA Research Service, Columbia, MO 65212.

cells can suppress development of autoimmune thyroiditis inrats.(13) However, myelin basic protein (MBP)-specific T celllines activated in vitro in the presence of TGF-b transferredmore severe EAE on transfer to recipient mice.(14) In addi-tion, autoimmune-prone MRL/lpr mice and systemic lupuserythematosus (SLE) patients reportedly have increased lev-els of endogenous TGF-b,(15,16) and endogenous TGF-b is im-portant for the development of spontaneous autoimmune thy-roiditis in mice.(17)

We previously hypothesized that G-EAT might result whenactivation of a T cell subset that negatively regulates EAT ef-fector cells is blocked, for example, by addition of anti-IL-2RmAb during in vitro activation of EAT effector cells.(4) As neu-tralization of IFN-g in culture or in recipient mice also pro-moted development of G-EAT,(5) we suggested that effectorcells for G-EAT may be Th2-like CD41 T cells or that activa-tion of G-EAT effector cells was inhibited by Th1-like cells.(4,5)

If so, addition of TGF-b during in vitro activation with MuTgmight promote the activation of G-EAT effector cells becauseTGF-b can suppress Th1 cells and cytokines.(6,7,18,19) The re-sults presented here show that spleen cells from MuTg-sensi-tized donors cultured with MuTg and TGF-b transfer moder-ately severe G-EAT to recipient mice. Importantly, the abilityof TGF-b to promote activation of G-EAT effector cells ap-parently does not require endogenous IL-12, as G-EAT inducedby cells activated in the presence of TGF-b was not affectedby neutralization of IL-12.

IL-12, a cytokine that promotes development of Th1 cellsand induces IFN-g production,(20) promotes the developmentof very severe G-EAT,(3) suggesting that all G-EAT effectorcells are probably not Th2-like. TGF-b can negatively regulateIL-12 production,(19–22) decrease IL-12Rb2 expression onCD41 T cells,(19) and inhibit the IFN-g-inducing effects of IL-12 on T cells and natural killer (NK) cells.(18,19,23) Because cellsactivated with MuTg and TGF-b induced G-EAT in recipientmice, TGF-b was added to cultures with IL-12 and MuTg insome experiments. TGF-b inhibited the IL-12-induced increasein IFN-g production and the effects of IL-12 on activation ofG-EAT effector cells. These opposing effects of TGF-b appar-ently involve separate pathways, one promoting activation ofG-EAT effector cells by an IL-12-independent pathway and anIL-12-dependent pathway that inhibits the IL-12-induced acti-vation of G-EAT effector cells.

MATERIALS AND METHODS

Cytokines and antibodies

Recombinant porcine TGF-b1 (rPoTGF-b1) (R & D Sys-tems, Minneapolis, MN) was activated by reconstituting 1mgTGF-b in 18 ml sterile 4 N HCl for 1 h at 4°C and adjustingto a concentration of 1 mg/ml with phosphate-buffered saline(PBS) containing 1% bovine serum albumin (BSA). Recombi-nant murine IL-12 (rMuIL-12) was obtained from R & D Sys-tems. Anti-TGF-b mAb (1.D11.16) was provided by Dr. WendyWaegell (Celtrix Pharmaceuticals, Santa Clara, CA).(10) Poly-clonal sheep anti-IL-12(3) was provided by Dr. Stanley Wolf(Genetics Institute, Cambridge, MA).

Mice

Female CBA/J mice, 6–10 weeks old, were obtained throughMr. Clarence Reeder (National Cancer Institute, Bethesda,MD). Donor mice were immunized twice at 10-day intervalswith 150 mg MuTg and 15 mg lipopolysaccharide (LPS) (Esch-erichia coli 011 B4, Sigma Chemical Co., St. Louis, MO).(2,3)

Adoptive transfer

Seven days after the second injection of MuTg and LPS,donor spleen cells were cultured at 1 3 107 cells/ml in RPMI1640 containing 25 mM HEPES buffer (Cell and Immunobiol-ogy Core Facility, University of Missouri), 5% fetal bovineserum (FBS) (Sigma Chemical Co.), 2 mM glutamine, MEMvitamin solution, nonessential amino acids, 1 mM sodium pyru-vate, and 5 3 1025 M 2-ME. Cells were cultured with 25 mg/mlMuTg. TGF-b (2–4 ng/ml), anti-TGF-b (5 mg/ml), anti-IL-12(5 mg/ml), and IL-12 (5 ng/ml) (or combinations) were addedas indicated for individual experiments. After 72 h, cells wereharvested and washed, and 3–3.5 3 107 cells were injected i.v.into irradiated (500 rad) recipient mice.(3,4) Recipient thyroidswere evaluated for EAT 19–21 days later, the time of maximalseverity of both lymphocytic EAT and G-EAT in this adoptivetransfer model.(3)

Evaluation of EAT

Thyroids were scored quantitatively for EAT severity (theextent of destruction of thyroid follicles) using a scale of 11

to 51 as previously described.(3) 11 EAT is defined as an in-filtrate of at least 125 cells in one or several foci, 21 indicates10–20 foci of cellular infiltration involving destruction of up to25% of the gland, 31 indicates that 25%–50% of the thyroid’snormal follicular structure is destroyed, 41 indicates that.50% of the gland is destroyed, and 51 indicates virtuallycomplete destruction of the thyroid, with few or no remainingfollicles. Thyroid lesions were also evaluated qualitatively.(3,4)

The thyroid infiltrate in lymphocytic EAT consisted primarilyof mononuclear cells with relatively few polymorphonuclearleukocytes (PMN). In this cell transfer model, lymphocytic EATlesions rarely exceeded 11 in severity. Thyroids designated Gin the tables had granulomatous histopathologic features char-acterized by proliferation and enlargement of thyroid follicularcells, with small to moderate numbers of neutrophils in the peri-follicular stroma, lymphocytes, histiocytes, epithelioid cells,and multinucleated giant cells.(3) The more severe (41–51)granulomatous lesions had intense inflammation with manyneutrophils, microabscess formation, necrosis, focal fibrosis,multinucleated giant cells, and numerous histiocytes. The gran-ulomatous inflammation in thyroids graded 41 to 51 charac-teristically extended beyond the thyroid to involve the adjacentconnective tissue and muscle.(3) Before scoring, slides werecoded and read by at least two of the investigators; differencesin interpretation were very rare.

ELISA assays

Serum levels of anti-MuTg IgG antibodies were measuredby ELISA using serum from individual mice as previously de-

FONDAL ET AL.972

scribed.(3) The contribution of particular IgG subclasses to thetotal IgG anti-MuTg response was assessed using alkaline phos-phatase-conjugated antibodies specific for mouse IgG1, IgG2a,and IgG2b.(3,24) Secondary antibodies were used at previouslydetermined optimal dilutions (1:6000 or 1:8000), titrated to givean OD ,0.05 with normal mouse serum (diluted 1:100) on anMuTg-coated plate or with a 1:100 dilution of anti-MuTg seraon plates coated with an irrelevant antigen (ovalbumin). Serialdilutions of sera were tested in duplicate. Supernatants from 72-h cultures were assayed for IFN-g by double sandwich ELISAassays as previously described.(3,24)

RT-PCR analysis

RNA isolation, RT-PCR amplification, and primer sequenceswere described in detail previously.(3,24,25) To determine the rel-ative initial amounts of target cDNA, each cDNA sample wasserially diluted, and each dilution was amplified with cytokine-specific primers. HPRT was used as a housekeeping gene toverify that equivalent amounts of RNA were amplified. PCRproducts were collected before the amplifications reached theplateau phase, separated by electrophoresis in 2% agarose gels,and visualized by ultraviolet (UV) light after ethidium bromidestaining. Densitometry analysis was performed using an IS-1000 Digital Imaging System (Life Sciences, Inc., St. Louis,MO). The resulting values were multiplied by 100 for generat-ing the values shown in Figure 2; that is, a ratio of cyto-kine/HPRT of 100 indicates a 1:1 relationship between thebands, and a ratio of 50 indicates the HPRT band is 2-fold higherthan the corresponding cytokine band. Dilutions of sampleswithin the linear relationship (usually 1:25 dilution) were usedfor analysis, and the densitometric units for each cytokine bandwere normalized to those obtained for the corresponding HPRTband. For comparison of relative levels of mRNA transcriptsbetween different samples, samples were reverse transcribedand amplified at the same time using aliquots of reagent fromthe same master mix.

Statistical analysis

A two-tailed Student’s t-test was used to compare differencesin disease severity and anti-MuTg responses between experi-mental groups. Values of p , 0.05 were considered to be sig-nificant.

RESULTS

Effects of TGF-b on T cell proliferation and EAT severity

To determine if TGF-b would modulate the in vitro activa-tion of EAT effector cells, spleen cells from MuTg-primeddonors were cultured 72 h with MuTg alone or with MuTg to-gether with TGF-b or anti-TGF-b. After 72 h, cells werewashed and transferred to recipient mice, and thyroids were re-moved 19–21 days later. In two representative experiments(Table 1), cells cultured with MuTg and 2 ng/ml TGF-b (1band 2b) transferred moderately severe (2–41) G-EAT, whichwas more severe (p , 0.005) and differed histologically fromthe 1–21 lymphocytic EAT induced by cells activated withMuTg alone (Table 1, 1a and 2a). Thyroids of recipients of cellsactivated with MuTg and TGF-b had follicular cell prolifera-tion and increased numbers of histiocytes, PMN, and multinu-cleated giant cells in addition to lymphocytes (Fig. 1G). TGF-b at 2 or 4 ng/ml also suppressed both MuTg-induced andconcanavalin A (ConA)-induced T cell proliferation (data notshown). Addition of anti-TGF-b together with MuTg and TGF-b neutralized the effects of TGF-b on EAT (Table 1, 1c) andT cell proliferation (data not shown). Addition of anti-TGF-bwith MuTg had no consistent effects on EAT severity (Table1, 1d and 2c), suggesting that endogenous TGF-b neither pro-moted nor suppressed activation of EAT effector cells.

For the experiments in Table 1, TGF-b was used at a con-centration of 2 ng/ml, which was shown in preliminary exper-iments to be optimal for activation of EAT effector cells. Cells

EFFECTS OF TGF-b ON EAT EFFECTOR CELL ACTIVATION 973

TABLE 1. EFFECT OF TGF-b AND ANTI-TGF-b ON IN VITRO ACTIVATION OF EAT EFFECTOR CELLS

IFN-gc Anti-MuTgd

MuTg TGF-b a-TGF-b 0 11 21 31 41 51 G (U/ml) IgG

1 2 2 1 3 1 0 0 0 1/5 7 0.254 6 0.081 1a1 1 2 0 0 2 2 1 0 5/5 1 0.450 6 0.042 1b1 1 1 0 3 3 0 0 0 0/6 6 0.222 6 0.013 1c1 2 1 0 1 3 1 0 0 1/5 3 0.502 6 0.022 1d1 2 2 2 6 1 0 0 0 0/9 170 0.382 6 0.057 2a1 1 2 0 1 1 5 1 0 6/8 60 0.598 6 0.078 2b1 2 1 1 6 2 0 0 0 0/9 120 0.665 6 0.039 2c

aCBA/J donors were immunized with MuTg and LPS as described in Materials and Methods. Spleen cells were cultured 72 hwith 25 mg/ml MuTg alone or with MuTg and TGF-b (2 ng/ml), anti-TGF-b (1.D11.16) (5 mg/ml), or both. Cells (3–3.5 3 107)were transferred to irradiated (500 rad) CBA/J recipients.

bNumber of mice with various degrees of severity of EAT 21 days after cell transfer. G, number of mice with granulomatouschanges in thyroids. p values are ,0.005 (1a vs. 1b and 2a vs. 2b). 1a–d and 2a–c are separate experiments.

cIFN-g production was determined by ELISA assay of supernatants from 72-h cultured cells.dMean anti-MutTg IgG 6 SEM OD410 of sera (1; 1600) from recipient mice. p values are ,0.05 (1a vs. 1b and 1d) and ,0.02

(2a vs. 2b and 2c).

Activationa Severity of EAT b

cultured with MuTg and 1 or 2 ng/ml TGF-b consistently trans-ferred G-EAT (usually 2–31 severity), whereas cells culturedwith MuTg and 4 ng/ml TGF-b (see Table 4, 1e vs. 1a) trans-ferred lymphocytic EAT (0–11), similar to cells cultured withMuTg alone. Cells cultured with TGF-b in the absence of MuTgdid not transfer EAT (data not shown).

IFN-g levels in 72-h supernatants from cells cultured withMuTg alone varied considerably between experiments and of-ten were very low. When IFN-g levels in supernatants wereclearly above background, (e.g., Table 1, 2a–c) TGF-b consis-tently decreased IFN-g production in 72-h culture supernatants.However, the decreased IFN-g production does not explain whycells activated in the presence of TGF-b induced G-EAT, asaddition of IL-12 to culture markedly increases both IFN-g pro-duction and the ability of cells to induce severe G-EAT.(3)

Although TGF-b inhibits the production of some classes ofIg by B cells,(6) it can also stimulate the production of partic-ular Ig isotypes, such as IgA and IgG2b.(6,26) Anti-MuTg re-sponses were generally higher in recipients of cells culturedwith MuTg plus TGF-b than in recipients of cells cultured withMuTg alone (Table 1). This was not due to an increase in anyparticular IgG subclass, as IgG1, IgG2a, and IgG2b were all in-creased (Tables 2, 3, and 4 and data not shown). MuTg-spe-cific IgA was not detected in any of the sera.

The experiments in Table 1 indicate that TGF-b promotesthe MuTg-induced activation of cells able to transfer moder-

ately severe G-EAT to recipient mice. Whether this involves adirect effect of TGF-b on the CD41 EAT effector cells or TGF-b acts indirectly through other cells present in the in vitro cul-ture is not known (see Discussion). To determine when TGF-b mediates these effects during the 72-h culture, TGF-b wasadded either at initiation of culture (0 h) or 24 or 48 h later(Table 2). Addition of TGF-b at culture initiation again causedthe transfer of moderately severe G-EAT. However, addition ofTGF-b at 24 or 48 h had no effect on the ability of cells to in-duce EAT in recipient mice, and anti-MuTg autoantibody re-sponses were generally decreased. These results indicate thatTGF-b must be present during the first 24 h of the 72-h cul-ture to promote activation of cells able to transfer G-EAT.

Induction of G-EAT by cells activated by MuTg andTGF-b is IL-12 independent

The ability of both IL-12 and anti-IL-2R mAb to induce acti-vation of cells able to transfer G-EAT is neutralized by the ad-dition of anti-IL-12 to culture.(3) To determine if endogenous IL-12 was required for the activation of G-EAT effector cells byTGF-b, anti-IL-12 was added to cultures with MuTg and IL-12or TGF-b (Table 3). As shown previously,(3) addition of anti-IL-12 prevented the ability of cells cultured with MuTg in the pres-ence of IL-12 to induce G-EAT and prevented the IL-12-inducedincrease in IFN-g production (Table 3, 1c vs. 1d). In marked con-

FONDAL ET AL.974

FIG. 1. Effect of TGF-b, TGF-b and anti-IL-12, IL-12, or both TGF-b and IL-12 on EAT induced by activation of donorspleen cells with MuTg. Hematoxylin and eosin (H & E)-stained thyroid sections from recipients of cells cultured with MuTgalone (A, F), MuTg and 2 ng/ml TGF-b (B, G), MuTg, 2 ng/ml TGF-b, and anti-IL-12 (C, H), IL-12 (D, I), or TGF-b and IL-12 (E, J). MuTg, TGF-b, and anti-IL-12 were added at day 0; IL-12 was added at 48 h of the 72-h culture. (A, F) 11 lympho-cytic EAT with mononuclear cell infiltrate. (B, C, G, H) 41 G-EAT showing granulomatous histopathologic features, includingfollicular cell proliferation, increased neutrophils and histiocytes in addition to lymphocytes. (D, I) 4–51 G-EAT with extensivegranulomatous changes, including increased neutrophils, microabscesses, and fibrosis. (E, J) 1–21 lymphocytic EAT withmononuclear cell infiltrate as in A and F. (A–E) 3 100. (F–J) 3 400.

trast, addition of anti-IL-12 with TGF-b increased the ability ofthese cells to transfer G-EAT (Table 3, 1a vs. 1b and 2a vs. 2b,and Fig. 1C,H). Thus, TGF-b apparently activates G-EAT ef-fector cells by an IL-12-independent pathway, and its effects areincreased when endogenous IL-12 is neutralized.

TGF-b interferes with the ability of IL-12 to inducesevere G-EAT

MuTg-sensitized donor spleen cells activated with MuTg inthe presence of exogenous IL-12 induce very severe G-EAT inrecipient mice, characterized by neutrophil infiltration, mi-croabscess formation, fibrosis, and a destructive inflammationof the thyroid.(3) These pathologic changes are much more pro-found than the granulomatous thyroid lesions induced by cellsactivated in the presence of TGF-b (Fig. 1D,I). Unlike TGF-b,IL-12 does not inhibit cell proliferation. It markedly increasesIFN-g production and mediates its effects primarily during thelast 24 h of the 72-h culture.(3) As TGF-b has been shown toinhibit the effects of IL-12(18,19,22) and IL-12 can decrease TGF-b production,(21,22) it was of interest to determine the effects ofadding both TGF-b and IL-12 on activation of EAT effectorcells. In the first experiment shown in Table 4 (1a–f), cells cul-tured with MuTg and TGF-b induced moderately severe G-EAT (31) (1b), whereas cells cultured with MuTg and IL-12added at culture initiation induced more severe (41) G-EAT(1c). Addition of both 2 ng/ml TGF-b and IL-12 to culture onday 0 suppressed the IL-12-induced increase in IFN-g produc-

tion and inhibited the effects of IL-12 on activation of G-EATeffector cells (Table 4, 1c vs. 1d). Increasing the amount ofTGF-b to 4 ng/ml resulted in even greater suppression of IL-12-induced G-EAT and suppression of anti-MuTg autoantibodyresponses (1f). It is important to note that although 2 ng/mlTGF-b added alone increased G-EAT severity in recipient mice(Table 4, 1a vs. 1b), 4 ng/ml TGF-b did not (1a vs. 1e). Thus,the ability of TGF-b to promote activation of G-EAT effectorcells is unrelated to its ability to suppress the IL-12-induced ac-tivation of G-EAT effector cells.

Addition of IL-12 as late as 48 h after culture initiation causesoptimal activation of effector cells to induce severe G-EAT.(3)

To determine when TGF-b was needed to suppress the IL-12-induced increase in effector cell function, IL-12 was added 48h after culture initiation, and TGF-b was added at either day 0or day 2 (Table 4, 2a–f). Because addition of IL-12 at 48 h doesnot result in increased production of IFN-g in 72-h culture su-pernatants,(3) IFN-g levels were not determined in this experi-ment. Cells cultured with MuTg and IL-12 added at 48 h in-duced severe G-EAT (4–51) in recipient mice (Table 4, 2c),whereas cells cultured with 2 ng/ml TGF-b added at day 0 in-duced less severe G-EAT (1–31) (Table 4, 2b vs. 2a). Addi-tion of TGF-b at 48 h did not activate cells to transfer G-EAT(Table 4, 2a vs. 2e). However, suppression of IL-12-inducedG-EAT was as great or greater when TGF-b was added at 48h as when it was added at time 0 (Table 4, 2c vs. 2d and 2e).Addition of TGF-b at 64 h (7–8 h before harvesting cells) hadno effect on IL-12-induced G-EAT (Table 5, A vs. D), and the

EFFECTS OF TGF-b ON EAT EFFECTOR CELL ACTIVATION 975

TABLE 2. TGF-b IS REQUIRED AT CULTURE INITIATION TO ENHANCE EAT

TGF-ba 0 11 21 31 41 51 G IgG1 IgG2b

— 1 4 0 0 0 0 0/5 0.415 6 0.102 0.147 6 0.0330 h 0 0 3 1 1 0 5/5 0.682 6 0.023 0.505 6 0.03824 h 3 2 0 0 0 0 0/5 0.103 6 0.047 0.015 6 0.00848 h 1 3 0 0 0 0 0/4 0.160 6 0.016 0.085 6 0.026

aSee footnote a, Table 1. MuTg was added to all cultures; TGF-b (2 ng/ml) was added at the indicated times after culture ini-tiation, and cells were harvested at 72 h.

bSee footnote b, Table 1. p values are ,0.005 (line 1 vs. line 2) and .0.1 (line 1 vs. line 3 or 4).cSee footnote d, Table 1. Serum dilutions were 1; 1600; secondary antibodies were anti-IgG1 or IgG2b.

Anti-MuTgcSeverity of EAT b

TABLE 3. TGF-b ACTIVATES G-EAT EFFECTOR CELLS BY AN IL-12-INDEPENDENT MECHANISM

IFN-gc

Anti-IL-12 TGF-b IL-12 0 11 21 31 41 51 G (U/ml) IgG1 IgG2b

1a 2 1 2 0 0 2 3 0 0 4/5 20 0.818 6 0.081 0.598 6 0.0701b 1 1 2 0 0 0 1 4 0 5/5 0 0.882 6 0.055 0.677 6 0.0661c 2 2 1 0 0 0 0 1 3 4/4 900 0.900 6 0.038 0.618 6 0.0591d 1 2 1 1 4 0 0 0 0 0/5 5 0.616 6 0.062 0.312 6 0.031

2a 2 1 2 0 0 4 1 0 0 0/5 9 0.899 6 0.082 0.599 6 0.1102b 1 1 2 0 0 0 1 4 0 5/5 6 1.014 6 0.008 0.721 6 0.042

a,b,c,dSee footnotes a, b, c, Table 1, and footnote c, Table 2. IL-12, anti-IL-12, and TGF-b were all added at day 0. IL-12 wasadded at 5 ng/ml, TGF-b at 2 ng/ml, and anti-IL-12 at 5 mg/ml. p values are ,0.005 (1a vs. 1b, 1c vs. 1d, and 2a vs. 2b. 1a–dand 2a–b represent separate experiments.

Activationa Anti-MuTgdSeverity of EATb

suppression of IL-12-induced G-EAT by TGF-b was reversedin the presence of 5 mg/ml anti-TGF-b added together withTGF-b at either time 0 or at 48 h (Table 5, C vs. E and B vs.F). This suggests that TGF-b interferes with the IL-12-inducedactivation of G-EAT effector cells at 48 h, the time when IL-12 is known to exert its enhancing effects on these effectorcells.(3) Moreover, the ability of TGF-b to suppress the effectsof IL-12 at 48 h is in marked contrast to the effects of TGF-balone, as TGF-b was required during the first 24 h of cultureto promote activation of G-EAT effector cells (Table 2). Takentogether, these results indicate that TGF-b promotes activationof G-EAT effector cells by a pathway that is distinct from thatthrough which TGF-b inhibits the IL-12-induced activation ofG-EAT effector cells.

Effects of TGF-b on production and expression of cytokines

A major function of IL-12 is to increase IFN-g productionby T cells and NK cells.(20) When IL-12 was added at the ini-

tiation of culture, IFN-g protein production was markedly in-creased(3) (Tables 3 and 4), and expression of IFN-g mRNAwas generally also increased(3) (Fig. 2B). Addition of anti-IL-12 reversed the effects of IL-12 on IFN-g mRNA expressionand protein production (Table 3 and Fig. 2). Addition of bothTGF-b and IL-12 suppressed the IL-12-induced increase inIFN-g protein production (Table 4) but had little or no effecton IFN-g mRNA expression (Fig. 2). Addition of TGF-b toculture generally decreased production of the low amounts ofIFN-g produced by MuTg-activated cells (Table 1) and also de-creased IFN-g mRNA expression (Fig. 2). Addition of TGF-bto cultures generally resulted in increased accumulation of IL-2 in 72-h culture supernatants (data not shown), probably be-cause the cells were proliferating minimally and thus not usingthe IL-2 they produced. There were no consistent effects ofTGF-b on IL-5 protein production (data not shown) or on IL-2 mRNA expression (Fig. 2).

Although IL-10 mRNA expression was decreased by TGF-b (Fig. 2), IL-10 protein production in 72-h culture supernatantswas very low and was not affected by TGF-b (1500 vs. 1130

FONDAL ET AL.976

TABLE 4. SUPPRESSION OF IL-12-INDUCED G-EAT BY TGF-b

IFN-g c

TGF-b IL-12 0 11 21 31 41 51 G (U/ml) IgG1 IgG2b

1a – – 2 3 0 0 0 0 0/5 19 0.437 6 0.083 0.258 6 0.0421b 0 h – 0 0 0 5 0 0 5/5 18 0.712 6 0.037 0.576 6 0.0271c – 0 h 0 0 0 1 4 0 5/5 1125 0.640 6 0.042 0.457 6 0.0631d 0 h 0 h 0 0 3 2 0 0 3/5 325 0.558 6 0.055 0.503 6 0.0361e 0 he – 1 4 0 0 0 0 0/5 7 0.335 6 0.074 0.215 6 0.0191f 0 he 0 h 5 0 0 0 0 0 0/5 300 0.355 6 0.088 0.219 6 0.058

2a – – 1 1 0 0 0 0 0/2 ND 0.423 6 0.138 0.297 6 0.1022b 0 h – 0 1 1 3 0 0 4/5 ND 0.537 6 0.125 0.449 6 0.0992c – 48 h 0 1 0 0 2 2 4/5 ND 0.573 6 0.056 0.476 6 0.0712d 0 h 48 h 0 0 0 5 0 0 5/5 ND 0.761 6 0.074 0.708 6 0.1302e 48 h – 1 4 0 0 0 0 0/5 ND 0.176 6 0.057 0.119 6 0.0352f 48 h 48 h 1 2 1 0 0 0 0/5 ND 0.340 6 0.067 0.194 6 0.050

a,b,c,dSee footnotes a, b, c, Table 1, and footnote c, Table 2. MuTg was added to all cultures. IL-12 was added at 0 h in ex-periment 1 (1a–f) and at 48 h in experiment 2 (2a–f). TGF-b was added 0 or 48 h after culture initiation as indicated.

eTGF-b was added at 4 ng/ml in 1e and 1f and at 2 ng/ml in all other groups. p values are ,0.005 (1a vs. 1b and 1c, 1c vs.1f, and 2c vs. 2f) and ,0.05 (1c vs. 1d, 2a vs. 2b, 2c, and 2d).

Activationa Anti-MuTgdSeverity of EAT b

TABLE 5. REVERSAL OF TGF-b-MEDIATED SUPPRESSION BY ANTI-TGF-b

Anti-MuTgc

Anti-TGF-b TGF-b IL-12 0 11 21 31 41 51 G IgG

A – – 1 0 0 1 3 1 0 5/5 0.647 6 0.135B – 0 h 1 0 4 1 0 0 0 1/5 0.795 6 0.045C – 48 h 1 0 2 3 0 0 0 0/5 0.302 6 0.082D – 64 h 1 0 0 0 2 3 0 4/5 1.071 6 0.094E 48 h 48 h 1 0 0 4 1 0 0 2/5 0.582 6 0.084F 0 h 0 h 1 0 0 0 2 3 0 4/5 0.981 6 0.040

a,b,cSee footnotes a, b, and d, Table 1. MuTg was added to all cultures. TGF-b (2 ng/ml) was added at 0, 48, or 64 h as indi-cated; IL-12 was added at 48 h (2.5 ng/ml). Anti-TGF-b was added at 0 or 48 h as indicated (5 mg/ml). p values are ,0.005 (Avs. B and C, B vs. F) and ,0.05 (C vs. E).

Activationa Severity of EAT b

pg/ml and 1140 vs. 1950 pg/ml in two different experiments).The apparent lack of effect of TGF-b on IL-10 protein pro-duction could be an artifact due to the very low amounts of IL-10 that were produced in these cultures. Although IL-4 proteinwas not detected in any culture supernatants (data not shown),IL-4 mRNA expression was consistently increased by additionof TGF-b to culture and was often further increased when anti-IL-12 was added with TGF-b (Fig. 2). Consistent with the in-creased IL-4 mRNA expression, expression of IL-12Rb2mRNA was reduced in cells cultured with TGF-b or TGF-bplus anti-IL-12, whereas addition of TGF-b together with IL-12 had relatively little effect on IL-12Rb2 or IL-4 mRNA ex-pression (Fig. 2). Although TGF-b mRNA expression wassometimes increased by addition of TGF-b to culture (Fig. 2),the increases in TGF-b mRNA expression were very modest,consistent with results of others, indicating that TGF-b may beregulated posttranscriptionally.(6) Although TGF-b has beenshown to decrease IL-12 production,(21,22) IL-12 p70 was un-detectable in culture supernatants except when exogenous IL-12 was added (data not shown). The alterations in cytokine pro-duction and mRNA expression shown here are consistent withthe idea that TGF-b alone acts on G-EAT effector cells via adifferent pathway than does IL-12, even though both cytokinescan promote activation of G-EAT effector cells. However, theseresults did not provide definitive information as to the mecha-nisms by which TGF-b inhibits the IL-12-induced activation ofG-EAT effector cells, as expression of IL-4 or IL-12Rb2mRNA expression was similar for cells cultured with IL-12 orwith IL-12 plus TGF-b (Fig. 2).

DISCUSSION

The results presented here indicate that MuTg-primed donorspleen cells activated with MuTg in the presence of TGF-b (2ng/ml) transfer moderately severe G-EAT to recipient mice,whereas cells activated with MuTg alone transfer a mild lym-phocytic form of EAT. This effect was dependent on the amountof TGF-b added to culture, as 1–2 ng/ml TGF-b activated cells

to induce G-EAT, whereas lower or higher amounts had no ef-fect (Table 4 and data shown). It is perhaps surprising that theamount of TGF-b apparently was so critical, although it wasreproducible in all experiments. The same amount (1–2 ng/ml)of TGF-b that was optimal for promoting activation of cells toinduce G-EAT was also found to be optimal for promoting ac-tivation of EAE effector cells.(14) The effects of TGF-b on EATeffector cells were reversed by anti-TGF-b mAb (Tables 1 and5). However, anti-TGF-b added alone with MuTg had no ef-fect on EAT (Table 1), suggesting that insufficient amounts ofendogenous TGF-b were present in culture to affect EAT sever-ity. TGF-b inhibited both MuTg-specific and ConA-inducedproliferation of cells in a dose-dependent manner (data notshown). Although TGF-b has been shown to inhibit T cell pro-liferation by regulating IL-2R expression,(6,27) there were nodetectable differences in IL-2R expression of cells cultured withMuTg in the presence or absence of TGF-b or IL-12 or both(data not shown).

The mechanism by which TGF-b promotes activation of cellsto induce G-EAT is not completely understood. TGF-b was re-quired during the first 24 h of the 72-h culture, as addition ofTGF-b at 24 or 48 h had no effect on EAT severity (Table 2).In studies in which T cells that induce EAE had increased ef-fector function after activation by antigen and TGF-b, TGF-binduced a shift in effector cell CD45R and CD62L expressiontoward a memory phenotype.(14) In our experiments, spleencells cultured 72 h with MuTg alone or MuTg plus TGF-b didnot differ in CD45R, CD44, or CD62L expression when ana-lyzed by flow cytometry (data not shown). However, as EATeffector cells comprise a minor portion of the total spleen cellpopulation, changes in expression of CD45R, CD44, CD62L,or IL-2R may have been present but would not have been de-tected in our experiments. In addition, the EAT effector cellspresumably had already differentiated to a memory phenotypeprior to culture. It is well established that the state of differen-tiation of T cells, for example, naive vs. memory cells, can in-fluence the effect of TGF-b on T cells.(6,7,28,29)

The induction of G-EAT by cells cultured with TGF-b andMuTg could be due to an indirect effect on effector CD41 T

EFFECTS OF TGF-b ON EAT EFFECTOR CELL ACTIVATION 977

FIG. 2. Cytokine gene mRNA expression by donor spleen cells cultured 72 h with MuTg alone or with MuTg and cytokines:2 ng/ml TGF-b added at day 0 or 5 ng/ml IL-12 added at 48 h (A) or at culture initiation (B), or anti-IL-12 (5 mg/ml at day 0),as indicated. cDNA was prepared from 72-h spleen cell cultures and amplified as described in Materials and Methods. Resultsare representative of four or five separate experiments and are expressed as the ratio of cytokine/HPRT (3100) of 1:25 cDNAdilutions. Each group represents pooled spleen cells from 72-h cultures (one sample per group).

A B

cells, as bulk populations of spleen cells were used for in vitroactivation. TGF-b inhibits NK cell proliferation, cytokine pro-duction, and NK-mediated cytotoxicity(21,30) and inhibits gen-eration of cytotoxic T lymphocytes (CTL) and IL-2-inducedlymphokine-activated killer (LAK) cells.(31) If NK cells orCD81 T cells negatively influence the activity of G-EAT ef-fector cells, TGF-b could inhibit their activation, resulting inincreased effector cell activity. Although CD81 T cells nega-tively regulate the activation of G-EAT effector cells,(32) bothenhancement of EAT effector function by TGF-b and TGF-b-mediated suppression of effector cell activation by IL-12 wereobserved when donor CD81 T cells were removed prior to cul-ture (data not shown). It is not known if depletion of NK cellswould influence the effects of TGF-b in this model, and an ef-fect of TGF-b on B cells that are required for transfer of G-EAT(33) cannot be excluded. TGF-b also could inhibit activa-tion-induced apoptosis of T cells,(34–36) resulting in increasedviability of EAT effector cells cultured in the presence of TGF-b. Although this possibility cannot be ruled out, there were noapparent differences in viability or cell recovery with any ofthe culture conditions used here.

The results of these experiments mimic in several ways theresults obtained when cells are activated with MuTg in the pres-ence of anti-IL-2R. Addition of the anti-IL-2Ra mAb M7/20during in vitro activation of effector cells with MuTg resultedin the transfer of moderately severe G-EAT to recipient mice.(4)

We previously hypothesized that anti-IL-2R mAb might pro-mote activation of G-EAT effector cells by blocking activationof cells that negatively regulate effector cell function,(4) as IL-2R1 T cells are known to negatively regulate autoimmune dis-eases.(37) As reported for the induction of G-EAT by cells cul-tured in the presence of anti-IL-2R,(4) TGF-b mediates itseffects during the first 24 h of culture (Table 2). TGF-b alsoinhibits cell proliferation, with increased accumulation of IL-2in culture supernatants. In contrast, when G-EAT effector cellsare activated by IL-12, cell proliferation is not inhibited, andIL-12 mediates its effects in this model primarily during the last24 h of the 72-h culture.(3) However, the ability of cells acti-vated in the presence of either anti-IL-2R or IL-12 to induceG-EAT is prevented by neutralization of endogenous IL-12,(3)

whereas anti-IL-12 has no effect on the ability of TGF-b to ac-tivate cells to transfer G-EAT (Table 3). Thus, the mechanismby which TGF-b promotes activation of cells that induce G-EAT is clearly distinct from that by which IL-12 and anti-IL-2R mAb promote activation of G-EAT effectors, as endoge-nous IL-12 is not required for TGF-b to activate cells to induceG-EAT. The ability of TGF-b together with IL-4 to induce ac-tivation of CD41 and CD81 T cells has been reported.(38,39) Ithas also been shown that memory IL-4-producing CD41 T cellsare not suppressed by TGF-b, whereas IFN-g-producing Th1memory cells are suppressed by TGF-b,(28) and TGF-b togetherwith IL-2 could promote expansion of IL-4-producing Th2cells.(36) Our results suggest that TGF-b could be acting throughan IL-4-dependent pathway in this model because addition ofTGF-b to spleen cell cultures resulted in increased IL-4 mRNAexpression (Fig. 2). Studies are in progress to determine if en-dogenous IL-4 is required for TGF-b to promote activation ofG-EAT effector cells.

The influence of TGF-b on EAT effector cells is likely tobe influenced both by the state of differentiation of the re-

sponding T cells and by other cytokines in culture.(6,28,29,38,39)

The influence of other cytokines was demonstrated here by theresults obtained when exogenous IL-12 was added to culturewith TGF-b and MuTg (Table 4). Effector cells activated withMuTg together with IL-12 induce very severe G-EAT.(3) How-ever, addition of TGF-b during activation of MuTg-primeddonor spleen cells with IL-12 and MuTg suppressed the effectsof IL-12, and these cells transferred less severe G-EAT (Table4). The mechanism by which TGF-b inhibits the IL-12-inducedactivation of G-EAT effector cells is apparently distinct fromthe mechanism by which addition of TGF-b alone promotes ac-tivation of these cells. Thus, the ability of TGF-b to promoteactivation of G-EAT effectors is IL-12 independent (Table 3)and is observed only when TGF-b is added at initiation of cul-ture (Table 2). In contrast, suppression of the effects of IL-12by TGF-b is optimal with amounts of TGF-b (4 ng/ml) that donot promote activation of G-EAT effectors. Moreover, TGF-bsuppresses IL-12-induced activation when added at 48 h (Table4), the time when IL-12 exerts its primary effects on EAT ef-fector cells(3) but when TGF-b added alone has no effect on ac-tivation of G-EAT effectors (Table 2). Thus, TGF-b apparentlymediates its effects in this model through two distinct pathways,an IL-12-independent pathway that results in cells transferringmoderately severe G-EAT when cells are cultured with MuTgand TGF-b, and an IL-12-dependent pathway resulting in thesuppression of severe G-EAT when cells are cultured withMuTg, TGF-b and IL-12. It is not surprising that TGF-b sup-pressed the IL-12-dependent pathway, as TGF-b can decreaseIL-12 production, and vice versa.(18,22,23,40–42) TGF-b can alsoinhibit the ability of cells to respond to IL-12(40,43) by inhibit-ing IL-12 signaling through the Jak-Stat pathway,(41) and asshown here (Fig. 2) and by others,(19,28,43) by inhibiting IL-12Rb2 expression. TGF-b also inhibited the IL-12-induced pro-duction of IFN-g (Table 4), as shown by others.(6,18,19,23,42,44)

The effects of TGF-b on G-EAT effector cells, however,probably are unrelated to its effects on IFN-g production be-cause activation of cells that induce G-EAT occurs both in thepresence of TGF-b, which inhibits IFN-g production, and inthe presence of IL-12, which promotes IFN-g production.(3) Inaddition, IFN-g2 /2 mice develop G-EAT,(45) cells cultured withanti-IFN-g induce G-EAT,(3,5) and IFN-g production in vitroor in vivo in recipient mice is unrelated to the ability of IL-12to activate cells to induce severe G-EAT.(3)

In summary, TGF-b has both positive and negative effectson the in vitro activation of effector cells for G-EAT. Thesestudies demonstrate that cells that induce a particular autoim-mune disease can be activated by more than one mechanism,as G-EAT can be induced by cells activated by both IL-12-de-pendent(3) and IL-12-independent (Table 3) pathways. The fi-nal outcome and severity of disease may depend on the inte-gration and balance of signals induced by different cytokinespresent at specific times during effector cell activation.

ACKNOWLEDGMENTS

This work was supported by NIH grant DK35527. W.F. andC.S. were supported by Gus T. Ridgel fellowships from theUniversity of Missouri. C.S. was also supported by National In-stitutes of Health Training grant T32 AI07276. We thank Patti

FONDAL ET AL.978

Mierzwa for excellent technical assistance, Karin Peterson forassistance with the cytokine ELISA assays, and Haiwen Tangand Robert Lopez for assistance with RT-PCR.

REFERENCES

1. CHARRREIRE, J. (1989). Immune mechanisms in autoimmunethyroiditis. Adv. Immunol. 46, 263–334.

2. BRALEY-MULLEN, H., JOHNSON, M., SHARP, G.C., andKYRIAKOS, M. (1985). Induction of autoimmune thyroiditis inmice with in vitro activated splenic T cells. Cell. Immunol. 93,132–143.

3. BRALEY-MULLEN, H., SHARP, G., TANG, H., KYRIAKOS,M., and BICKEL, J.T. (1998). Interleukin 12 promotes activationof effector cells that induce a severe destructive granulomatousform of experimental autoimmune thyroiditis. Am. J. Pathol. 152,1347–1358.

4. BRALEY-MULLEN, H., SHARP, G.C., BICKEL, J.T., and KYR-IAKOS, M. (1991). Induction of severe granulomatous experi-mental autoimmune thyroiditis in mice by effector cells activatedin the presence of anti-interleukin 2 receptor antibody. J. Exp. Med.173, 899–912.

5. STULL, S.J., SHARP, G.C., M. KYRIAKOS, M., BICKEL, J.T.,and BRALEY-MULLEN, H. (1992). Induction of granulomatous ex-perimental autoimmune thyroiditis in mice with in vitro activated ef-fector T cells and anti IFN-g antibody. J. Immunol. 149, 2219–2226.

6. LETTERIO, J.J., and ROBERTS, A.B. (1998). Regulation of im-mune responses by TGF-BETA. Annu. Rev. Immunol. 16, 137–161.

7. PRUD’HOMME, G.J., and PICCIRILLO, C.A. (2000). The in-hibitory effects of transforming growth factor-beta-1 (TGFb1) inautoimmune diseases. J. Autoimmunity 14, 23–42.

8. RACKE, M.K., DHIB-JALBUT, S., CANNELLA, B., ALBERT,P.S., RAINE, C.S., and McFARLIN, D.E. (1991). Prevention andtreatment of chronic relapsing experimental allergic encephalo-myelitis by transforming growth factor b1. J. Immunol. 146,3012–3017.

9. JOHNS, L.D., FLANDERS, K.C., RANGES, G.E., and SRIRAM,S. (1991). Successful treatment of experimental allergic en-cephalomyelitis with transforming growth factor b1. J. Immunol.147, 1872–1879.

10. BRANDES, M.E., ALLEN, J., OGAWA, Y., and WAHL, S.M.(1991). Transforming growth factor beta suppresses acute andchronic arthritis in experimental animals. J. Clin. Invest. 87,1108–1113.

11. KING, C., DAVIES, J., MUELLER, R., LEE, M.S., KRAHL, T.,YEUNG, B., O’CONNOR, E., and SARVETNICK, N. (1998).TGFb1 alters APC preference, polarizing islet antigen responsive-ness toward a Th2 phenotype. Immunity 8, 601–613.

12. PICCIRILLO, C.A., CHANG, Y., and PRUD’HOMME, G.J.(1998). TGF-b1 somatic gene therapy prevents autoimmune dis-ease in nonobese diabetic mice. J. Immunol. 161, 3950–3956.

13. SEDDON, B., and MASON, D. (1999). Regulatory T cells in thecontrol of autoimmunity: the essential role of transforming growthfactor beta and interleukin 4 in the prevention of autoimmune thy-roiditis in rats by peripheral CD41CD45RC2 cells andCD41CD82 thymocytes. J. Exp. Med. 189, 279–288.

14. WEINBERG, A.D., WHITHAM, R., SWAIN, S.L., MORRISON,W.J., WYRICK, G., HOY, C., VANDENBARK, A.A., andOFFNER, H. (1992). Transforming growth factor b enhances thein vivo effector function and memory phenotype of antigen spe-cific T helper cells in experimental autoimmune encephalomyelitis.J. Immunol. 148, 2109–2117.

15. LOWRANCE, J.H., O’SULLIVAN, F.X., CAVER, T.E.,WAEGELL, W., and GRESHAM, H.D. (1994). Spontaneous elab-

oration of transforming growth factor b suppresses host defenseagainst bacterial infection in autoimmune MRL/lpr Mice. J. Exp.Med. 180, 1693–1703.

16. CAVER, T.E., O’SULLIVAN, F.X., GOLD, L.I., and GRESHAM,H.D. (1996). Intracellular demonstration of active TGFb1 in B cellsand plasma cells of autoimmune mice. IgG bound TGF-b1 sup-presses neutrophil function and host defense against Staphylococ-cus aureus infection. J. Clin. Invest. 98, 2496–2506.

17. BRALEY-MULLEN, H., CHEN, K., WEI, Y., and YU, S. (2001).Role of TGFb in development of spontaneous autoimmune thy-roiditis in NOD.H-2h4 mice. J. Immunol. (in press).

18. SCHMITT, E., HOEHN, P., HUELS, C., GOEDERT, S., PALM,N., RUDE, E., and GERMANN, T. (1994). T helper type 1 de-velopment of naïve CD41 T cells requires the coordinate action ofinterleukin-12 and interferon-gamma and is inhibited by trans-forming growth factor-beta. Eur. J. Immunol. 24, 793–798.

19. KITANI, A., FUSS, I.J., NAKAMURA, K., SCHWARTZ, O.M.,USUI, T., and STROBER, W. (2000). Treatment of experimental(trinitrobenzene sulfonic acid) colitis by intranasal administrationof transforming growth factor (TGFb1) plasmid: TGF-b1-medi-ated suppression of T helper cell type 1 response occurs by inter-leukin-10 induction and IL-12 receptor b2 chain downregulation.J. Exp. Med. 192, 41–52.

20. TRINCHIERI, G. (1995). Interleukin-12: a proinflammatory cyto-kine with immunoregulatory functions that bridge innate resistanceand antigen-specific adaptive immunity. Annu. Rev. Immunol. 13,251–276.

21. PARDOUX, C., ASSELIN-PATUREL, C., CHEHIMI, J., GAY,F., MAMI-CHOUAIB, F., and CHOUAIB, S. (1997). Functionalinteraction between TGFb and IL-12 in human primary allogeneiccytotoxicity and proliferative response. J. Immunol. 158, 136–143.

22. MARTH, T., STROBER, W., SEDER, R.A., and KELSALL, B.L.(1997). Regulation of transforming growth factor-beta productionby interleukin-12. Eur. J. Immunol. 27, 1213–1220.

23. BELLONE, G., ASTE-AMEZAGA, M., TRINCHIERI, G., andRODECK, U. (1995). Regulation of NK cell function by TGF b.J. Immunol. 155, 1066–1073.

24. PETERSON, K.E., SHARP, G.C., TANG, H., and BRALEY-MULLEN, H. (1999). B7-2 has opposing roles during the activa-tion versus effector stages of experimental autoimmune thyroidi-tis. J. Immunol. 162, 1859–1867.

25. TANG, H., and BRALEY-MULLEN, H. (1997). Intravenous ad-ministration of deaggregated mouse thyroglobulin suppresses in-duction of experimental autoimmune thyroiditis and expression ofboth Th1 and Th2 cytokines. Int. Immunol. 9, 679–687.

26. SNAPPER, C.M., WAEGELL, W., BEERNINK, H., and DASCH,J.R. (1993). Transforming growth factor-b1 is required for secre-tion of IgG of all subclasses by LPS-activated murine B cells invitro. J. Immunol. 151, 4625–4636.

27. KEHRL, J.H., WAKEFIELD, L.M., ROBERTS, A.B.,JAKOWLEW, S., ALVEREZ-MON, M., DERYNCK, R., SPORN,M.B., and FAUCI, A.S. (1986). Production of transforming growthfactor b by human T lymphocytes and its potential role in the reg-ulation of T cell growth. J. Exp. Med. 163, 1037–1050.

28. LUDVIKSSON, B.R., SEEGERS, D., RESNICK, A.S., andSTROBER, W. (2000). The effect of TGF-b1 on immune responsesof naïve versus memory CD41 Th1/Th2 cells. Eur. J. Immunol.30, 2101–2111.

29. SEDER, R.A., MARTH, T., WIEVE, M.C., STROBER, W., LET-TERIO, J.J., ROBERTS, A.B., and KELSALL, B. (1998). Factorsinvolved in the differentiation of TGF-b-producing cells from naïveCD41 T cells: IL-4 and IFN-g have opposing effects, while TGF-b regulates its own production. J. Immunol. 160, 5719–5728.

30. ROOK, A.H., KEHRL, J.H., WAKEFIELD, L.M., ROBERTS,A.B., SPORN, M.B., BURLINGTON, D.B., LANE, H.C., andFAUCI, A.S. (1986). Effects of transforming growth factor b on

EFFECTS OF TGF-b ON EAT EFFECTOR CELL ACTIVATION 979

the functions of natural killer cells: depressed cytolytic activity andblunting of interferon responsiveness. J. Immunol. 136, 3916–3920.

31. FONTANA, A., FREI, K., BODMER, S., HOFER, E., SCHREIER,M.H., PALLADINO, M., Jr., and ZINKERNAGEL, R.M. (1989).Transforming growth factor-b inhibits the generation of cytotoxicT cells in virus-infected mice. J. Immunol. 143, 3230–3236.

32. BRALEY-MULLEN, H., McMURRARY, R.W., SHARP, G.C.,and KYRIAKOS, M. (1992). Regulation of the induction and res-olution of granulomatous experimental autoimmune thyroiditis inmice by CD81 T cells. Cell. Immunol. 153, 492–504.

33. BRALEY-MULLEN, H., SHARP, G.C., and KYRIAKOS, M.(1994). Differential requirement for autoantibody-producing Bcells for induction of lymphocytic versus granulomatous experi-mental autoimmune thyroiditis. J. Immunol. 152, 307–314.

34. CERWENKA, A., KOVAR, H., MAJDIC, O., and HOLTER, W.(1996). Fas- and activation-induced apoptosis are reduced in hu-man T cells preactivated in the presence of TGF-b1. J. Immunol.156, 459–464.

35. GENESTIER, L., KASIBHATLA, S., BRUNNER, T., andGREEN, D.R. (1999). Transforming growth factor b1 inhibits Fasligand expression and subsequent activation-induced cell death inT cells by downregulation of c-myc. J. Exp. Med. 189, 231–240.

36. ZHANG, X., GIANGRECO, L., BROOME, H.E., DARGAN, C.M.,and SWAIN, S.L. (1995). Control of CD4 effector fate: transform-ing growth factor b1 and interleukin 2 synergize to prevent apopto-sis and promote effector expansion. J. Exp. Med. 182, 699–709.

37. SHEVACH, E.M. (2000). Regulatory T cells in autoimmunity.Annu. Rev. Immunol. 18, 423–450.

38. LINGNAU, K., HOEHN, P., KERDINE, S., KOELSCH, S., NEU-DOERFL, C., PALM, N., RUEDE, E., and SCHMITT, E. (1998).IL-4 in combination with TGF-b favors an alternative pathway ofTh1 development independent of IL-12. J. Immunol. 161,4709–4718.

39. ERARD, F., GARCIA-SANZ, J.A., MORIGGL, R., and WILD,M. (1999). Presence or absence of TGFb determines IL-4-inducedgeneration of type 1 or type 2 CD8 T cell subsets. J. Immunol. 162,209–214.

40. MARTH, T., STROBER, W., and KELSALL, B.L. (1996). Highdose oral tolerance in ovalbumin TCR-transgenic mice. Systemicneutralization of IL-12 augments TGF b secretion and T cell apop-tosis. J. Immunol. 157, 2348–2357.

41. BRIGHT, J.J., and SRIRAM, S. (1998). TGFb inhibits IL-12-in-duced activation of Jak-Stat pathway in T lymphocytes. J. Im-munol. 161, 1772–1777.

42. TAKEUCHI, M., ALARD, P., and STREILEIN, J.W. (1998). TGF-b promotes immune deviation by altering accessory signals of anti-gen-presenting cells. J. Immunol. 160, 1589–1597.

43. GORHAM, J.D., GULER, M.L., FENOGLIO, D., GUBLER, U.,and MURPHY, K.M. (1998). Low dose TGFb attenuates IL-12 re-sponsiveness in murine Th cells. J. Immunol. 161, 1664–1670.

44. HUNTER, C.A., BERMUDEZ, L., BEERNINK, H., WAEGELL,W., and REMINGTON, J.S. (1995). Transforming growth factorb inhibits interleukin-12 induced production of interferon-g by nat-ural killer cells: a role for transforming growth factor b in the reg-ulation of T cell-independent resistance to Toxoplasma gondii. Eur.J. Immunol. 25, 994–999.

45. TANG, H., SHARP, G.C., PETERSON, K.E., and BRALEY-MULLEN, H. (1998). IFN-g-deficient mice develop severe gran-ulomatous experimental autoimmune thyroiditis (EAT) witheosinophil infiltration in thyroids. J. Immunol. 160, 5105–5112.

Address reprint requests to:Dr. Helen Mullen

Division of Immunology, Department of MedicineUniversity of Missouri School of Medicine

Room M450 Medical SciencesColumbia, MO 65212

Tel: (573) 882-7150Fax: (573) 882-1380

E-mail: mullenh@health.missouri.edu

Received 12 April 2001/Accepted 2 July 2001

FONDAL ET AL.980