Role of interleukin-4 in down-regulation of contact sensitivity by γδ T cells from tolerized T-cell receptor α^−/− mice

Abstract

Contact sensitivity (CS) is a classical example of an in vivo T-cell-mediated immune response that is under regulation. Such down-regulation can be mediated by αβ T cells in mice that are tolerized by prior exposure to high doses of antigen. In contrast, we demonstrated previously that such high-dose antigen tolerance in T-cell receptor (TCR) α^−/− H-2^d mice induced antigen-specific, apparently major histocompatibility complex-unrestricted, CD4⁻ CD8⁻γδ T cells, that also could down-regulate CS responses antigen-specifically in vivo, and also inhibited in vitro production of IFN-γ. In the present experiments we employed H-2^b-deficient TCRα^−/− and TCRβ^−/− mice, owing to different molecular constructs than were used previously, and confirmed that tolerized γδ T cells in these different H-2^bαβ TCR^−/− mice down-regulated CS. Thus, γδ T-cell suppressor function was not limited to mice bearing a special transgenic TCRα^−/− DNA construct. Furthermore, employing monoclonal antibody and complement depletion in vitro and adoptive transfer in vivo, characterized the phenotype of these γδ down-regulatory T cells as: CD3⁺, CD28⁺, CD40-ligand⁺, Fas⁺, FcγR⁺ and NK1.1⁻. Also, in vitro antigen desensitization of these trinitrophenyl (TNP)-specific TCRγδ⁺ down-regulatory cells was achieved with soluble TNP-bovine serum albumin (BSA), but not with oxazolone-BSA, showing that these suppressive γδ T cells have antigen-specific receptors. Moreover, employing monoclonal antibody blocking of γδ suppressors in vitro, and of recipients in vivo, we showed that interleukin-4 (IL-4) was involved in this down-regulation of CS by γδ T cells, while IL-10 and transforming growth factor-β₂ were not. In summary, generation of antigen-specific, double-negative, γδ suppressor cells, by tolerance of high antigen doses in TCRα^−/− mice, appears to be a general phenomenon, and IL-4 production is involved in their down-regulation of the T helper type 1 cells that mediate CS.

INTRODUCTION

Cutaneous delayed-type hypersensitivity (DTH) responses, and related contact sensitivity (CS) reactions, are classical examples of in vivo cell-mediated immune responses mediated by αβ T-cell receptor-positive (TCR⁺) CD4⁺, antigen-major histocompatibility complex (MHC) class II-specific T helper type 1 (Th1) lymphocytes.¹ The elicitation of CS reactions is under strict regulation. This includes positive regulation by antigen-specific αβ TCR⁺ up-regulatory cells.² CS reactions also are under the positive influence of antigen-non-specific γδ TCR⁺-assisting T cells, that normally are present in the spleen, and can be activated to circulate, and appear to express Vγ5 and Vδ4 TCR preferentially.⁵

Besides such positive regulation, CS responses are also negatively regulated by antigen-specific αβ TCR⁺ CD8⁺ down-regulatory suppressive T cells.⁷ In addition, we also showed previously that some CS responses are regulated negatively by antigen-specific, apparently MHC-unrestricted, γδ TCR⁺ T cells. These γδ TCR⁺ suppressive cells were induced by intravenous injection of high doses of antigen into TCRα^−/− mice. Similar high antigen dose tolerogenesis in normal TCRα^+/+ mice, instead induced down-regulatory cells that were αβ TCR⁺. The down-regulatory γδ T cells were shown to be double negative (CD4⁻ CD8⁻). Strikingly, quite low numbers of these suppressive γδ⁺ T cells (about 2×10³ adoptively transferred per recipient) could systemically mediate this down-regulation. In summary, regulation of CS responses can be mediated by up-regulatory² and down-regulatory⁷αβ T cells, and also up-regulatory, and down-regulatory γδ T cells.⁸

The current work focuses on further characterization of the γδ TCR⁺ down-regulatory cells that are induced by high antigen dose in TCRα^−/− mice. The data describe a more complete surface phenotype, and confirm antigen-specificity by an assay involving surface antigen receptors. We also describe the role of interleukin-4 (IL-4) in the mechanism of action of down-regulatory γδ T cells, which suppress CS responses, that are mediated by Th1 CS-effector T cells.

MATERIALS AND METHODS

Mice

BALB/c (H-2^d) female mice, 5−7 weeks old, were obtained from Jackson Laboratories, Bar Harbor, ME. Homozygous TCRα knock-out (TCRα^−/−) mice⁹ on an H-2^d background were bred and supplied locally through the courtesy of Dr Adrian Hayday. In some experiments, separately derived C57BL/6 (H-2^b) TCRα^−/− or TCRβ^−/− mice, also on an H-2^b background, were used and were obtained from Jackson Laboratories.¹⁰ All mice were maintained in microisolator cages, changed in a laminar flow hood, and were fed autoclaved food and water. MHC haplotypes were determined by fluorescence-activated cell sorter (FACS; Becton Dickinson FACSTAR PLUS, Mountain View, CA) of peripheral blood lymphocytes with anti-MHC-class I^b [phycoerythrin (PE)-conjugated], and anti-MHC class I^d monoclonal antibody (mAb) [fluorescein isothiocyanate (FITC)-conjugated; Pharmingen, San Diego, CA]. The locally bred and supplied TCRα^−/− mice^9,11 were tested by Southern analysis for the absence of normal TCRα DNA, and the presence of the neomycin-resistance gene contained within TCRα DNA.⁹ Each experimental group consisted of four or five mice.

Reagents

Picryl chloride (PCl) [trinitrophenyl (TNP) chloride; Chemica Alta, Edmonton, Alberta, Canada], was recrystallized from methanol/H₂O and protected from light and humidity; trinitrobenzene sulphonic acid (TNBSA, a water soluble analogue of PCl; Eastman Chemicals, Rochester, NY); oxazolone (OX; Gallard Schlesinger, Carle Place NY), bovine serum albumin (BSA; Sigma, St. Louis, MO); were obtained from the manufacturers. TNP- BSA and OX-BSA were prepared as described previously.¹²

Cytokines, antibodies and complement

The following mAb were used: anti-CD28 (37.51), anti-CD40L (MR1), anti-Fas (Jo2), anti-FcγR (2·4G2), anti-CD3 (145-2C11) and anti-NK1.1 (PK136), all obtained from Pharmingen, San Diego, CA, and pan anti-TCRδ (clone; UC7-135 D5) hybridoma supernatants, containing no less than 10 μg of mAb per ml, through the kindness of Dr Jeffrey Bluestone. The mAb to transforming growth factor-β₂ [TGF-β₂; clone; 1D11.16, mouse immunoglobulin G1 (IgG1)]¹³ was kindly provided by Dr Wendy Waegel of Celtrix Labs, Santa Clara, CA. The mAb to murine IL-4 (clone; 11B11, rat IgG1) was kindly provided by Dr Robert Coffman at DNAX Research Institute, Palo Alto, CA, and also by the US Government under a contract with the National Cancer Institute (NCI) and the National Institutes of Health. The mAb to IL-10 (clone; JES 5-2A5.1.1, rat IgG1) also was a kind gift of Dr Robert Coffman. Control rat and mouse IgG were from Sigma, St. Louis, MO, and low-tox rabbit complement (C′) was from Pel-Freeze, Brown Deer, WI.

Adoptive cell transfer of CS

Donors of immune cells were actively contact sensitized by topical application of 0·15 ml of 5% PCl in a 1:3 acetone: ethanol mixture, to the shaved abdomen, chest and hind feet on day 0. On day 4, lymph nodes and spleens were harvested and a single cell suspension was prepared. Then, a mixture of 7×10⁷ immune spleen and lymph node cells were transferred adoptively intravenously (i.v.) into naive syngeneic recipients. Immediately after the transfers, recipients were challenged on each ear with 20 μl 0·8% PCl in olive oil:acetone, 1:1. Subsequent increase in ear swelling was determined 24 hr later using a micrometer (Mitutoyo, Paramus, NJ), and expressed in units of swelling of 0·01 mm±SE. Each experiment consisted of a group of non-immune mice that only were challenged on the ears with 0·8% PCl, and their background ear swelling (≈2 units at 24 hr), was subtracted from the responses of the experimental groups, to yield the net ear swelling responses that are shown in the Figures.

Intravenous induction of down-regulatory cells

To induce high antigen dose tolerance, naive mice were injected i.v. twice with 3 mg TNBSA (1% TNBSA in distilled H₂O, readjusted to pH 7·2 with 1 m NaOH) on days 0 and 3. Then on day 7, spleen cells from these tolerized mice were harvested as a source of putative down-regulatory cells.

Down-regulation assay by mixing of CS effector cells with γδ down-regulatory cells

PCl immune cells (7×10⁷), from contact sensitized α^+/+ mice, were incubated in vitro for 30 min at 37° with 5×10⁷ spleen cells (containing unseparated regulatory T cells),⁸ from i.v. TNBSA-tolerized TCRα^−/− mice, or from tolerized TCRβ^−/−, or from tolerized TCRα^+/+ mice. The cell mixture was washed and transferred i.v. to normal recipients which were then skin tested immediately for adoptive transfer of CS reactivity.

To determine the phenotype of the γδ down-regulatory cells, 5×10⁷ spleen cells from i.v. TNBSA-tolerized TCRα^−/− mice first were incubated in vitro with 10 ml of UC7 hybridoma supernatant, or separate purified mAb to surface determinants [50 μg of mAb in 10 ml of phosphate-buffered saline (PBS) containing 2% fetal calf serum (FCS)], for 40 min on ice and then were incubated with C′ (1:75 dilution) for 45 min at 37°. Each step was followed by two cell washes in PBS supplemented with 2% FCS. Tolerized cells treated identically with C′ alone served as controls. Thereafter, 5×10⁷ cell equivalents of regulatory cells that had been treated with mAb and C′, or with C′ alone, were incubated in vitro with untreated 7×10⁷ 4-day PCl-immune CS effector cells, for 30 min at 37°. Then the cell mixture was injected i.v. into naive, syngeneic recipients that were challenged immediately on the ears with 0·8% PCl, and then tested for CS ear responses. Immune CS-effector cells (7×10⁷) injected alone served as a positive control. The size of active and adoptive CS reactions to PCl is haplotype-dependent, and in H-2^d mice is much higher than in H-2^b animals.

In vitro antigen treatment (desensitization) of γδ down-regulatory T cells with homologous and heterologous hapten-conjugated ligands, prior to the cell mixing and transfer assay

Spleen cells from TNBSA-tolerized TCRα^−/− mice were incubated at 1×10⁸ cells in a 10-ml volume, for 1 hr at 37°, with 100 μg/ml of either BSA (control group), or TNP-BSA or OX-BSA (experimental groups). Then, 5×10⁷ regulatory cells preincubated with BSA, or TNP-BSA, or OX-BSA, were washed, mixed with 7×10⁷ 4-day PCl-immune CS effector cells, and incubated together for 30 min at 37° in PBS with 2% FCS, then washed, and then the mixture was transferred adoptively i.v. into naive syngeneic recipients, that were tested for CS, as described above.

Treatment of down-regulatory cells and recipients with anticytokine mAbs

In some experiments, 5×10⁷ spleen cells from i.v. TNBSA-tolerized TCRα^−/− (H-2^d) mice were preincubated with 500 μg of the following mAb: anti-IL-4, anti-IL-10, anti-TGF-β₂, control mouse IgG or rat IgG, or PBS, for 20 min on ice. Then 4-day PCl-immune CS-effector cells were added to each group, and the cell mixtures were incubated for 30 min at 37°. In addition to these experimental groups, there also was a positive control group containing CS-immune effector cells alone (without regulatory cells), that were treated just with PBS alone or treated with 500 μg of anti-IL-4 mAb for 30 min at 37° to judge the effect of anti-IL-4 mAb on the CS cell transfers. Cell mixtures were washed and transferred i.v. into naive recipients, or into mice treated with 500 μg of each corresponding anticytokine mAb (see above), to maintain anticytokine mAb treatment in vivo, or were treated with control mouse IgG or rat IgG, or PBS, one day before transfer. After cell transfers, recipient mice were challenged on the ears with PCl and tested for elicitation of CS.

Statistics

Double-tailed Student’s t-test was used to assess the significance of differences between groups, with P < 0·05 taken as a minimum level of significance.

RESULTS

Expanded phenotypic determination of down-regulatory γδ T cells

To determine the phenotype of these γδ down-regulatory T cells more precisely, we tested spleen cells from TNBSA-tolerized TCRα^−/− (H-2^d) mice, in which down-regulation of Th1 CS responses was shown previously to be due to contained γδ⁺ CD4⁻ CD8⁻ T cells.⁸ Separate aliquots of tolerized down-regulatory spleen cells were treated with different mAbs and C′ and then were incubated individually with 7×10⁷ BALB/c indicator 4-day PCl-immune CS-effector cells for 30 min at 37°. After adoptive i.v. cell transfer of the mixed cells to the naive recipients, they were tested subsequently for elicitation CS.

The results in Fig. 1 show that spleen cells from TNBSA-tolerized TCRα^−/− mice when mixed with CS effector cells inhibited the adoptive transfer of CS (compare positive Group A to regulated Group B). The cell population in the TNBSA-tolerized TCRα^−/− cells, that mediated this down-regulatory phenomenon, expressed γδ TCR (Group C), confirming our previous results.⁸ This down-regulatory function was also reversed by treatment with anti-CD3 mAb (Group D), anti-CD28 mAb (Group E), or anti-CD40L mAb (Group F), or anti-Fas mAb (Group G), or by anti-FcγR mAb (Group H), but not by anti-NK1.1 mAb (Group I).

Figure 1.

Phenotype of down-regulatory cells from tolerized α^−/− mice; 5×10⁷ spleen cells from TNBSA-tolerized TCR α^−/− (H-2^d) mice were treated in vitro with the following mAbs: anti-γδ (UC7) (Group C), anti-CD3 (Group D), anti-CD28 (Group E), anti-CD40L (Group F), anti-Fas (Group G), anti-FcγR (Group H) and anti-NK1.1 (Group I); followed by incubation with diluted C′ at 37°. Group B down-regulatory cell transfer controls were treated with PBS+C′ alone. Then these mAb-treated down-regulatory cells, were incubated with 7×10⁷ syngeneic PCl-immune CS effector cells, and then were transferred together i.v. into naive syngeneic recipients. Mice were challenged immediately on the ears with PCl and the ear swelling responses were determined 24 hr later. As a positive control, PCl-immune CS-effector cells alone also were transferred i.v. (Group A). Statistical significance: Group B versus Group A, P < 0·001; Group C versus Group B, P < 0·02; Group D versus Group B, P < 0·05; Groups E and F versus Group B, P < 0·001; and Groups G and H versus Group B, P < 0·01.

Thus, taken together with our prior results,⁸ the down-regulatory cells were: TCRγδ⁺, CD3⁺, CD4⁻, CD8⁻, CD28⁺, CD40L⁺, FcγR⁺, Fas⁺NK1.1⁻, and MHC-unrestricted T cells. However, these functional phenotype experiments may have depleted the targeted cells participating in down-regulation, both in vitro, via complement, and in vivo via the reticuloendothelial system, and thus could not distinguish whether the positive markers were themselves involved in down-regulatory activity, nor whether the various determined marker antigens were present on the same cells in the mixed population.

High antigen dose tolerogenesis treatment also induced down-regulatory T cells, in other TCRα^−/− mice, and in TCRβ^−/− mice

We also attempted to induce high antigen dose unresponsiveness in TCRβ^−/− and TCRα^−/− (H-2^b) mice that originated from another source,⁹ that was different than the previously used TCRα^−/− (H-2^d) mice,^8,9,11 and employed different DNA constructs to achieve deletion of TCRα-chain expression. PCl-immune CS effector cells from C57BL/6 mice were mixed and incubated with 5×10⁷ spleen cells from tolerized C57BL/6 TCRα^+/+ control mice, or from C57BL/6 background TCRα^−/−, or TCRβ^−/− mice. Mixtures of tolerized cells and CS-effector cells were transferred adoptively into naive syngeneic normal C57BL/6 recipients that were then tested immediately by PCl ear challenge to elicit CS.

Figure 2 shows that treatment of both of the C57BL/6 background TCRα^−/− and TCRβ^−/− mice (Groups C and D), with high doses of TNBSA antigen i.v., induced down-regulatory cells that inhibited CS transferred by PCl-immune CS-effector T cells; similar to the effect observed with tolerized control TCRα^+/+ mice (Group B). Therefore, these data confirmed that the previously observed cell transferrable down-regulation in tolerized TCRα^−/− (H-2^d) mice, was not limited to a select group of TCR^−/− mice bearing a special TCRα DNA construct, or to some artifact of their particular genetic manipulation. Also, since the TCRα^−/− mice still might theoretically have a small subpopulation of active TCRββ cells,¹⁴ the confirmatory results in TCRβ^−/− mice, in which TCRα^+/+ nor TCRββ cells do not occur, further confirm that the down-regulatory cells have γδ TCR.

Figure 2.

Induction of γδ down-regulatory cells in mice with different genetic constructs used to eliminate αβ TCR⁺ cells. As donors of down-regulatory cells, we used TCRβ^−/− mice, or TCRα^−/− mice, on a strictly H2^b (C57BL/6) background, and originating from a different laboratory than the previously tolerized TCRα^−/− mice (H-2^d). Spleen cells (5×10⁷) from i.v. TNBSA-tolerized C57BL/6 TCRα^−/− mice (Group C), or from TCRβ^−/− mice (Group D), or from TCRα^+/+ control mice (Group B), were incubated with 7×10⁷ PCl-immune CS-effector cells from normal TCRα^+/+ C57BL/6 mice. Positive control, CS effector cells from normal TCRα^+/+ C57BL/6 mice were mixed with PBS. Then the cell mixtures were transferred into naive syngeneic recipients that were tested immediately on the ears for PCl ear swelling CS responses. Statistical significance: Group B versus Group A, P < 0·05; Group C versus Group A, P < 0·02; and Group D versus Group A, P < 0·02.

Antigen-specific desensitization of the γδ down-regulatory T cells

To determine whether the down-regulatory function of tolerized γδ T cells could be inhibited by incubation in vitro with appropriate antigen in solution, 1×10⁸ spleen cells from TNBSA-tolerized TCRα^−/− (H-2^d) mice were incubated in vitro with 100 μg/ml of either control antigen BSA, or TNP-BSA or OX-BSA to test for their antigen specificity. Then, 5×10⁷ of these antigen-treated γδ down-regulatory cells were incubated with 7×10⁷ TNP-specific 4-day immune CS-effector T cells, and then transferred together i.v. to naive recipients that were challenged immediately on the ears with PCl to test for CS elicitation.

Figure 3 shows that preincubation of the TNBSA-induced γδ down-regulatory T cells with TNP-BSA (Group C), but not OX-BSA (Group D), nor plain BSA (Group B), inhibited their ability to suppress the CS effector T cells (Group A). Thus, γδ down-regulatory cells seemed to have TNP-antigen-specific receptors on their surface, that when engaged by soluble specific antigen hapten−protein conjugates, inhibited their down-regulatory function.

Figure 3.

Antigen specificity of γδ down-regulation. Spleen cells (10⁸) from TNBSA-tolerized TCRα^−/− (H-2^d) mice were incubated with either 100 μg/ml BSA as a control (Group B), or with TNP-BSA (Group C), or with OX-BSA (Group D), and then washed thoroughly.Treated down-regulatory cells from each resultant group were incubated together with 7×10⁷ PCl-immune CS-effector cells, and then the cell mixtures were transferred i.v. into naive syngeneic recipients. For positive controls, 7×10⁷ PCl-immune CS effector cells alone were adoptively transferred i.v. (Group A). Then, all recipients were PCl ear challenged, and tested for CS elicitation. Statistical significance: Group A versus Group B, P < 0·02; Group C versus Group B, P < 0·05; and Group D versus Group A, P < 0·01.

IL-4, but neither IL-10 nor TGF-β₂ is responsible for regulatory cell inhibition of the target CS effector cells

We performed experiments to determine a possible mechanism of down-regulation mediated by the tolerized γδ T cells, that perhaps involved cytokines. To begin anticytokine blocking, spleen cells (5×10⁷) from tolerized TCRα^−/− mice (H-2^d) were preincubated with 500 μg of the following mAb: either anti-IL-4, anti-IL-10, anti-TGF-β₂, or a mixture of all of these anticytokine mAb (anti-IL-4, anti-IL-10 and anti-TGF-β₂), or control mouse or rat IgG, or PBS, as described in the Materials and Methods. Then, the mAb-treated cells were mixed with CS effector cells, washed and transferred together i.v. into either naive recipients, or into mice treated intraperitoneally with 500 μg of each of the above listed anticytokine mAb, to continue the anticytokine block, or with a mixture of all the employed anticytokine mAb, or with the control antibodies described above.

Figure 4 shows that treatment with anti-IL-4 inhibited the down-regulatory γδ cell-mediated inhibitory function of CS-effector cells (Group D). This anti-IL-4 blocking was done both in vivo and in vitro and even a low dose of 50 μg anti-IL-4 was effective (data not shown). In contrast, treatment in vivo and in vitro with anti-IL-10 (Group F), or with anti-TGF-β₂ mAb (Group G), produced results similar to control antibodies (Groups H and I), and thus did not reverse suppression by the down-regulatory γδ T cells. Additionally, immune cells alone were transferred into naive mice as positive controls (Group A), or into recipients treated with 500 μg of 11B11 anti-IL-4 mAb to judge the effect of anti-IL-4 mAb on CS cell transfers, which had no effect on CS transfers (Group B).

Figure 4.

Anti-IL-4 mAb blocks the suppressive activity of γδ down-regulatory cells from tolerized α^−/− mice. Spleen cell aliquots from i.v. TNBSA-tolerized TCRα^−/− (H-2^d) mice were preincubated in vitro with 500 μg of the following mAb: anti-IL-4 (Group D), or anti-IL-10 (Group F), or anti-TGF-β₂ (Group G), or a mixture of all three anticytokine mAb (Group E), or with control mouse IgG (Group H), or rat IgG (Group I), or with PBS alone (Group C). Additionally, in Group B, to test the effect of anti-IL-4 mAb on the CS transferring T cells, the PCl-immune CS effector cells themselves were preincubated with 500 μg of anti-IL-4 mAb, and then were transferred into mice that also were treated with 500 μg 11B11 anti-IL-4 mAb. Then the resultant groups of cells were incubated further in vitro with 7×10⁷ PCl-immune CS-effector cells, and then transferred together into naive recipients (Group A), or into recipient mice treated in vivo with the same anticytokine mAb as above (Group D, anti-IL-4; Group E, mixed anti-IL-4 and anti-IL-10 and anti-TGF-β₂; Group F, anti-IL-10, and Group G, anti-TGF-β₂) (see the Materials and Methods for details), or recipients treated with control IgG (Group H, mouse IgG; Group I, rat IgG). Positive controls (Group A) just received PCl-immune CS-effector cells. Recipients then were tested for CS elicitation by topical ear challenge with PCl. Statistical significance: Group B versus Group A, not-significant; Group C versus Group A, P < 0·01; Group D versus Group C, P < 0·01; Group E versus Group C, P < 0·001; Group F versus Group A, P < 0·01; and Group G versus Group A, P < 0·001.

We concluded that of the known inhibitory cytokines examined, IL-4 was principally responsible for suppression of Th1 CS-effector cells by the γδ down-regulatory cells.

DISCUSSION

Our previous study showed that TCRα^−/− H-2^d mice when injected i.v. with high doses of antigen developed antigen-specific, double-negative (CD4⁻ CD8⁻), and apparently MHC-unrestricted γδ down-regulatory T cells that effectively inhibit elicitation of Th1-dependent CS responses, in an adoptive cell transfer system.⁸ Data presented in studies by others also clearly demonstrate that down-regulatory γδ T cells, are also operative in numerous immunobiological systems; including: oral tolerance,¹⁵ autoimmunity,¹⁶ transplantation,¹⁷ respiratory allergy to aerosolized antigen,¹⁸ tumour immunity¹⁹ and pathogenesis of infectious diseases.²⁰

In our previous study, we employed separation with anti-TCRδ mAb-coated magnetic beads to show that the down-regulatory T cells induced by soluble i.v. antigen in TCRα^−/− mice express γδ TCR.⁸ In the current study we confirmed these previous data by using another technique, namely, in vitro depletion of mAb-coated target cells with rabbit C′, and then via i.v. transfer in vivo, where the reticuloendothelial system probably also contributed to depletion of mAb- and C′-coated cells.

Combined with our prior results,⁸ we show the suppressor cells are: TCRγδ⁺, CD3⁺, CD4⁻, CD8⁻, CD28⁺, CD40L⁺, FcγR⁺, Fas⁺ and NK1.1⁻ cells. Expression of the CD28 molecule by TCRγδ down-regulatory cells suggests they, like classical TCRαβ⁺ cells, may require costimulation, perhaps via B7.1 (CD80) or B7.2 (CD86) on antigen-presenting cells to become fully activated, but here in an MHC-unrestricted fashion. However, studies of others also show that some γδ T cells alternatively can be costimulated by CD43, or by Thy-1 molecules to provide a CD28-independent activation pathway in CD28-deficient mice.²¹

CD40L is usually expressed by activated CD4⁺αβ T cells, and plays an important role in B-cell activation,²² and in some instances for IgE production via CD40L on γδ T cells.²³ Other findings suggest that CD40L stimulation could contribute to expansion and functional maturation of γδ T cells.²⁴ Therefore expression of CD40L by γδ down-regulatory T cells of CS following activation by i.v. antigen, may be associated with their expansion, maturation, or function, particularly in TCRα^−/− mice, that produce these regulatory cells more readily, compared to TCRα^+/+ mice⁸.

Our study shows that γδ down-regulatory T cells express Fas. This Fas cell surface molecule (APO-1, CD95) can transduce a signal leading to physiological apoptotic death of cells that bear it, when it is engaged by corresponding ligand (Fas-L) on other T cells.²⁵ The role of Fas and Fas-L in the immune system seems to be cytotoxicity or regulation.²⁶ In our case we can speculate that binding Fas could in some instances be a form of positive regulation, to eliminate the down-regulatory γδ T cells, and thus balance immune responses. At present, it is under investigation whether the γδ T cells may possibly suppress via cytotoxicity directed against the CD4⁺αβ TCR⁺ CS-effector cells.

The current study suggests FcγR expression by γδ down-regulatory T cells. FcγR usually are associated with leucocytes, macrophages and B cells; but there are some examples of FcR on T cells.²⁷ Prior studies showed that resting γδ T cells isolated from spleen do not express FcγR, but their activation induced membrane FcR for IgM, IgA and for IgG.²⁸ This further suggests that the i.v. antigen induced down-regulatory γδ T cells in TCRα^−/− mice are activated, similar to other reports that FcR⁺ T cells, were involved in immunoregulation.²⁷

There are increasing reports that some NK1.1⁺ T cells, particularly those producing IL-4, may play a role in immunoregulation.²⁹ Moreover it was shown that NK1.1⁺ cells account for 10% of CD4⁻ CD8⁻ TCRγδ⁺ thymocytes, and that γδ T cells with natural killer (NK) cell markers represent a significant portion of γδ T cells in the circulation of humans.³⁰ In our experiments we found that the down-regulatory γδ T cells were negative for expression of NK1.1 antigen, ruling out a role for NK1.1⁺γδ T cells in this system.

The potential antigen and MHC specificity of γδ T cells in general is controversial, as is also the case for down-regulatory function by γδ T cells. However, prior studies suggest that γδ down-regulatory cells can be antigen-specific as found also in our studies.^8,15,18,31 Our present data showed that the function of γδ down-regulatory T cells induced by a high i.v. dose of reactive TNP hapten could be blocked completely by incubation in vitro with related TNP-BSA conjugates, but not by the unrelated OX-BSA hapten protein complex, or by BSA alone. Therefore, these data confirmed our previous finding that these particular γδ down-regulatory T cells function in an antigen-specific manner⁸ and suggest that in this system they can directly recognize unprocessed antigen, and thus may be hapten specific, like antibodies, and some T cells.³² Indeed, DNA sequence studies of antigen-combing site CDR3 regions of some γδ TCR have suggested more similarities to antibodies, rather then to analogous sequences of peptide/MHC binding αβ TCR.³³ Moreover, it is consistent with this antibody-like hypothesis for γδ TCR antigen-specificity, that our previous study showed that the i.v. TNBSA induced γδ down-regulatory T cells were apparently MHC unrestricted.⁸ Thus peptide/MHC binding, usual of αβ T cells, may be less important in this instance.

The down-regulatory effect of tolerized γδ T cells on CS effector T cells that produce Th1 cytokines, such as interferon-γ (IFN-γ),⁸ might in part be due to a direct or indirect inhibition of IFN-γ production by the CS-effector cells, as our prior in vitro results suggested.⁸ The mechanism for such IFN-γ inhibition might be mediated by release of inhibitory Th2 cytokines, such as IL-4, IL-10, or possibly TGF-β₂, by the down-regulatory γδ T cells. To test this hypothesis we tried to block inhibitory cytokines both in vitro and then in vivo, by using appropriate anticytokine mAbs. Our data show that initial pretreatment of down-regulatory γδ T cells in vitro, and subsequently in vivo, with anti-IL-10 or anti-TGF-β₂, did not reverse suppression. In contrast, anti-IL-4, even at a low dose of 50 μg per mouse, was uniformly inhibitory of down-regulation. Production of IL-4 by these down-regulatory γδ T cells would be consistent with the reported capacity of γδ T cells from TCR α^−/− mice to support B cells to produce copious amounts of various antibodies,⁴ and the demonstration that many primary γδ T-cell clones express Th2 cytokines, and function as Th2 cells, by producing IL-4.³⁴ Thus, down-regulation of CS by γδ T cells may be another instance for Th2 cytokine regulation or diversion of effector Th1 cells that in this system are the CS-effector T cells. However, the methods we employed did not allow determination of whether IL-4 was produced directly by the γδ down-regulatory cells, or indirectly via other cells, such as macrophages.³⁵

In summary, DTH and CS are important examples of in vivo Th1-mediated effector responses that in some instances can be crucial in defence against microbial pathogens or tumour cells, and can serve as effector cells in some allergic and autoimmune diseases. This widespread biological relevance of Th1 cells producing either beneficial or deleterious effects, may be the need for such tight regulation of these various responses. It is thus possible that changes in this immunoregulation, that sometimes involves regulatory γδ T cells, may result in decreased immune resistance to serious infections or cancers, if down-regulation increases, or perhaps to abnormally increased unwanted immune responses, that are observed in allergies and autoimmunity, if γδ T-cell down-regulation were decreased. Further studies to examine these questions are being undertaken.