gammadelta T-cell clones from intestinal intraepithelial lymphocytes inhibit development of CTL responses ex vivo

Abstract

Oral administration of antigen induces a state of tolerance that is associated with activation of CD8+ T cells that can transfer unresponsiveness to naïve syngeneic hosts. These T cells are not lytic, but they inhibit development of antibody, CD4+ T helper cell, and CD8+ cytotoxic T lymphocyte (CTL) responses upon adoptive transfer into naïve, syngeneic mice. In addition, we have shown that depletion of gammadelta T cells by injection of the anti-delta chain antibody (GL3) down modulates the expression of gammadelta T-cell receptor (TCR) and inhibits the induction of oral tolerance to ovalbumin. Oral administration of antigen also fails to induce tolerance in TCR delta-chain knockout mice suggesting that gammadelta T cells play a critical, active role in tolerance induced by orally administered antigen. To further study the contribution of gammadelta T cells to tolerance, murine gammadelta T cells were isolated from intraepithelial lymphocytes (IEL) of the small intestine by stimulation with splenic filler cells, concanavalin A and growth factors. gammadelta IEL lines demonstrated lytic activity in a redirected lysis assay. gammadelta T-cell clones express different gammadelta TCR genes and secrete large amounts of interleukin (IL)-10, but little or no IL-2, IL-4, or interferon-gamma. gammadelta IEL clones expressed transforming growth factor-beta1 and macrophage migration inhibitory factor, as well as IL-10, mRNA. Moreover, gammadelta T-cell clones potently inhibited the generation of CTL responses by secreted molecules rather than by direct cell-to-cell contact.

Figure 1

Growth requirements of a γδ IEL line. The γδ IEL4.B2 line was incubated at 10⁵ cells/ml alone or with 10⁶ irradiated spleen cells/ml, 2·5 µg/ml Con A, or 2 ng/ml rIL-2 as indicated above. The cells were harvested 3–4 days later, washed, viable cells counted and the growth index (cell number at end of culture/input cell number) recorded. Growth index on day 0 reflects the growth of the input cells, which were subdivided and grown under the indicated conditions, during the previous cycle.

Figure 2

Cell-surface antigens expressed by IEL T cell lines. IEL4.A4, a representative γδ IEL line, and IEL16-2, a representative αβ IEL line, were incubated with FITC-labelled isotype control (the left-most curve in each histogram) or antibodies that are indicated in the above flow cytometry histograms.

Figure 3

Cell surface antigens expressed by γδ T-cell clones. Two representative γδ IEL subclones (IEL23-4·23·14 and IEL4.A4.8.25) were treated with FITC-labelled antibodies shown on the x-axis and PE-labelled antibodies on the y-axis. Appropriately labelled isotype controls were used to set the quadrants for the negative controls.

Figure 4

Identification of γ and δ chains expressed by two IEL clones. The functional γ- and δ-chains for two independently derived, γδ IEL subclones were cloned and sequenced. The nucleotide sequences and the translated amino acids are depicted above.

Figure 5

Lytic activity of IEL T cells. The lytic activity of various T cells was measured at various effector : target ratios by incubation with ⁵¹Cr-labelled P815 plus or minus 1 μg/ml anti-CD3 antibody. The activity of an OVA-specific CTL line and the indicated γδ and αβ IEL lines are shown in (a) while that of clones IEL23-4.23.14 (IEL23) and IEL4.A4.8.25 (IEL4) are compared to the αβ 16-2 line in (c). The lytic activity of the γδ IEL22/N4 line was measured in the presence or absence of 1·8 mm CaCl₂ (b).

Figure 6

Requirements for secretion of IL-10 by γδ IEL clones. γδ clone, IEL4.A4.1, was incubated at 2·5 × 10⁵ cells/ml with or without 2·5 × 10⁶/ml irradiated (2000 rad) spleen cells from the indicated strains of mice plus or minus 2·5 μg/ml Con A (a). γδ IEL4.A4.8.25 and γδ IEL23-4.23.14 were incubated at 10⁵ cells/ml alone or with 20 U/ml IL-2 and/or IL-4 (b). Supernatants were collected after 24 h and assayed for IL-10 by ELISA. (Note that IEL4.A4.1 and IEL4.A4.8 are sister clones from the same bulk line; IEL4.A4.8.25 is a subclone of IEL4.A4.8 and IEL23-4·23.14 is a subclone of clone IEL23-4·23.)

Figure 7

Expression of cytokine mRNA by γδ and αβ IEL. The indicated riboprobes were hybridized to 10 μg RNA isolated from αβ IEL4.A4.8.25 and IEL23-4.23.14 and αβ IEL16.2, digested with RNAse, and resolved by polyacrylamide gel electrophoresis. Riboprobes for the housekeeping genes, L32 or GAPDH, were used to verify loading of equivalent amounts of RNA onto the gel.

Figure 8

Immunosuppressive activity of representative IEL clones. The suppressive activity was assessed by adding the indicated numbers of γδ IEL23-4.23.14 (a, left panel) or αβ IEL16.2 (a, right panel) to triplicate wells of 96-well plate containing 2·5 × 10⁵ B6 responder cells and 10 × 10⁵ irradiated BALB/c stimulator cells in a total volume of 200 μl. Tenfold limiting dilutions of γδ IEL23-4·23·14 (b, left panel) or αβ IEL16·2 (b, right panel) were added to 12 or 24 replicate wells of MLR cultures as described in (a); each symbol represents a replicate well. After 5 days of incubation, ⁵¹Cr-labelled P815 targets were added to each well. There were no statistical differences by Student's t-test between any groups except the asterisk indicating P < 0·02 compared to the control group containing no IEL.

Figure 9

γδ IEL suppression in transwell cultures. Lymph node cells (6 × 10⁶/ml) from B6 mice primed with OVA in CFA were added to 24-well plates. γδ IEL23-4.23.14 or αβ IEL16.2 cells (2 × 10²) were added directly to the wells or to inserts that separated the IEL from the responder cells by a membrane with 0·4 μm pores. Cultures were stimulated with 20 μg/ml OVA and assayed for CTL activity with ⁵¹Cr-labelled E.G7-OVA targets 5 days later.