Multiple Levels of Activation of Murine CD8+ Intraepithelial Lymphocytes Defined by OX40 (CD134) Expression: Effects on Cell-Mediated Cytotoxicity, IFN-γ, and IL-10 Regulation

Heuy-Ching Wang and John R. Klein
J Immunol December 15, 2001, 167 (12) 6717-6723; DOI: https://doi.org/10.4049/jimmunol.167.12.6717

Abstract

The involvement of OX40 (CD134) in the activation of CD8+ intestinal intraepithelial lymphocytes (IELs) has been studied using freshly isolated IELs and in vitro CD3-stimulated IELs. Although freshly isolated CD8+ IELs exhibited properties of activated T cells (CD69 expression and ex vivo cytotoxicity), virtually all CD8+ IELs from normal mice were devoid of other activation-associated properties, including a lack of expression of OX40 and the ligand for OX40 (OX40L) and an absence of intracellular IFN-γ staining. However, OX40 and OX40L expression were rapidly up-regulated on CD8 IELs following CD3 stimulation, indicating that both markers on IELs reflect activation-dependent events. Unlike IELs, activated lymph node T cells did not express OX40L, thus indicating that OX40-OX40L communication in the intestinal epithelium is part of a novel CD8 network. Functionally, OX40 expression was exclusively associated with IELs with active intracellular IFN-γ synthesis and markedly enhanced cell-mediated cytotoxicity. However, OX40 costimulation during CD3-mediated activation significantly suppressed IL-10 synthesis by IELs, whereas blockade of OX40-OX40L by anti-OX40L mAb markedly increased IL-10 production. These findings indicate that: 1) resident CD69+OX40− IELs constitute a population of partially activated T cells poised for rapid delivery of effector activity, 2) OX40 and OX40L expression defines IELs that have undergone recent immune activation, 3) OX40+ IELs are significantly more efficient CTL than are OX40− IELs, and 4) the local OX40/OX40L system plays a critical role in regulating the magnitude of cytokine responses in the gut epithelium.

The activation status of intestinal intraepithelial lymphocytes (IELs)3 continues to be a topic of controversy. Because of their strategic location at mucosal surfaces, IELs are considered to play an important role in the protection against the entry and dissemination of pathogens and foreign Ags. Moreover, because of extensive Ag exposure in mucosal tissues, delivery of local effector activity likely requires a process whereby activation is achieved rapidly and efficiently; in fact, there is evidence that this is the case. For example, small intestine human and murine CD8+ IELs exhibit phenotypic and functional properties characteristic of activated effector cells such CD69 expression (1) and a capacity to lyse target cells in ex vivo assays when effector and target cells are ligated via the TCR or CD3ε (2, 3, 4, 5, 6). That the latter is a true property of murine IELs is supported by multiple lines of evidence. First, ex vivo cytolytic activity of IELs in short-term 2- to 4-h cultures has been independently reported by a number of laboratories (2, 3, 4, 5, 6). Second, this activity cannot be attributed to cell isolation methodologies given that peripheral splenic or lymph node T cells exposed to the IEL extraction protocol do not acquire cytotoxic activity or CD69 expression, and because IELs isolated by mechanical disruption without enzyme digestion retain cytotoxic activity (our unpublished observations). Third, in situ activation of IELs is dependent upon the presence of luminal Ags as seen in studies in which germ-free mice express little lytic activity, whereas mice housed under conventional conditions are cytotoxic (6), thus suggesting that the intestinal microenvironment conditions the functional response of resident lymphocytes.

OX40 was first identified as an activation molecule on rat lymphocytes (7). An equivalent molecule for mouse T cells has been recently described and, although originally linked to the activation of CD4 T cells, it is now clear that mouse CD8 T cells activated either by CD3 cross-linking or by lectin stimulation also express OX40 (8, 9, 10). The ligand for OX40 (OX40L) is expressed at varying levels on B cells, T cells, and APCs including dendritic cells (DCs) (11, 12, 13). Although OX40L can be constitutively expressed at low levels on those cells, its expression is increased during activation (14, 15), thus accounting for the costimulatory properties of OX40L-mediated lymphocyte stimulation (12, 13, 16, 17).

Recently, OX40 has been implicated in human intestinal inflammatory responses as seen by increases in OX40 expression in the intestine of patients with inflammatory bowel disease (18) and in patients with active celiac disease, Crohn’s disease, and ulcerative colitis (19). However, those studies did not establish whether OX40 expression is a function-associated characteristic of cells undergoing local activation thus directly linking OX40 to the inflammatory process itself, or whether expression is acquired as a secondary or ancillary event of inflammation. Circumstantial evidence favoring the former comes from studies of graft-vs-host disease in mice in which OX40 expression is up-regulated on T cells in secondary lymphoid tissues before the onset of intestinal tissue destruction (20). At present, little else is known about the expression of OX40 and IEL activation in mice.

To better understand the stages involved in IEL activation, we have examined the expression of OX40 and OX40L on IELs in freshly isolated cell preparations and in in vitro CD3-stimulated cultures. These findings indicate that unlike T cells in other peripheral immune compartments, small intestine IELs exist in differential states of activation, and they define a mechanistic role for OX40 in the generation and regulation of local immunity. A model for OX40 involvement in intestinal immunity and immunopathology is presented.

Materials and Methods

Mice

Adult female BALB/c and C57BL/6 mice, 8–12 wk of age, were purchased from Harlan Sprague-Dawley (Houston, TX) and were maintained at the University of Texas (Houston, TX) vivarium. Animals were housed and used under conditions approved by the University of Texas Animal Welfare Committee.

Lymphoid cell isolations

Small intestine tissues were removed and Peyer’s patches were dissected out. Tissues were flushed of fecal material, opened longitudinally, and cut into 3- to 4-mm pieces in RPMI 1640 supplemented with FCS (10% v/v), 100 U/ml penicillin-streptomycin, 2 mM l-glutamine, and 5 × 10−5 M 2-ME (all from Sigma-Aldrich, St. Louis, MO). Tissue fragments were rinsed several times in Ca2+/Mg2+-free PBS and stirred at 37°C for 30 min in Ca2+/Mg2+-free PBS containing 5 mM EDTA and 2 mM DTT (Sigma-Aldrich). Cells were filtered successively through three 10-cc syringe barrels containing wetted nylon wool, centrifuged, suspended in 3 ml of 40% isotonic Percoll (Sigma-Aldrich), layered on top of 70% isotonic Percoll, and centrifuged for 20 min at 600 × g. IELs were recovered from the Percoll interface. In some cases, the Percoll step was repeated. Based on expression of the leukocyte common Ag (LCA), CD45, by flow cytometry, 85–92% of the cells were IELs, with >98% viability by trypan blue exclusion. Mesenteric lymph node cells were obtained by pressing tissues through a 60-mesh stainless steel screen.

Abs and flow cytometry

Abs used in this study were purified anti-CD3 (145-2C11), FITC-anti-CD8α (53-6.7), FITC-anti-CD8β (53-5.8), FITC-anti-CD45 LCA (30-F11), PE-anti-CD69 (H1.2F3), biotin-anti-CD134/OX40 (OX86), biotin-anti-OX40L (RM134L), purified anti-OX40L (RM134L (NA/LE)), anti-hamster (G94-56), purified hamster Ig, purified rat IgG (R35-95), propidium iodide, streptavidin-CyChrome, anti-CD16/32 Fc block (2.4G2) (all from BD PharMingen, San Diego, CA); and anti-rat IgG/M (312-005-044) (Jackson ImmunoResearch Laboratories, West Grove, PA). Intracellular IFN-γ staining was done using a commercial cell staining kit (BD PharMingen) with the manufacturer’s reagents, protocols, and controls. Stained cells were analyzed on a FACSCalibur flow cytometer using CellQuest software (BD Biosciences, San Jose, CA). Background staining by species-matched control reagents is demarcated by the position of cells in the lower left histogram quadrant.

Cell culture

For flow cytometric analyses, purified IELs were cultured at a density of 1.2–2.0 × 106 cells/2 ml in 24-well plates coated with anti-CD3 mAb or in uncoated or control mAb-coated wells with supplemented RPMI 1640 (as described above) in the presence or absence of rIL-2 and/or rIL-15 (Sigma-Aldrich). In some initial experiments, wells were coated overnight by direct binding of anti-CD3 (145-2C11) mAb or isotype-matched hamster control mAb to plates. However, in most experiments, CD3 stimulation was done by coating wells overnight at 4°C with 10 μg/ml anti-hamster mAb, washed with PBS, and incubated for 1 h at 37°C with mAb 145-2C11. Unbound Ab was removed by washing before the addition of cells. For CD3/OX40 costimulation, wells were coated overnight at 4°C with a suspension of 10 μg/ml anti-hamster mAb and 10 μg/ml anti-rat mAb. Wells were washed with PBS and incubated for 1 h at 37°C with 5 μg/ml mAb 145-2C11 and 5 μg/ml anti-OX40 mAb. For soluble blockade using anti-OX40L mAb, wells were coated overnight at 4°C with 10 μg/ml anti-hamster mAb, washed with PBS, incubated for 1 h at 37°C with 5 μg/ml mAb 145-2C11, and washed, and cells were added with the indicated amount of soluble anti-OX40L mAb. Cells were collected after 24 or 48 h, stained, and analyzed by flow cytometry after gating onto viable cells using propidium iodide exclusion, or supernatants were recovered for cytokine assays. For determination of cytokine-mediated cell survival, viability was scored at 24 and 48 h by trypan blue exclusion. For in vitro stimulation of lymph node cells, 15 × 106 mesenteric lymph node cells were cultured in 5 ml of supplemented RPMI 1640 in 60-mm petri dishes coated with anti-hamster mAb followed by mAb 145-2C11 as described above.

Cytotoxicity and cytokine assays

Redirected cytotoxicity assays were done as previously reported (3), except that Fc-receptor-bearing C26 tumor target cells were used. For analyses of redirected cytotoxicity from CD3-stimulated in vitro cultures, IELs were collected and dead cells were removed by centrifugation through a 40%/70% Percoll gradient. Cells at the interface were collected, washed, and separated into OX40+ and OX40− populations by magnetic-activated cell sorting (Miltenyi Biotec, Auburn, CA). In brief, IELs were reacted for 10 min at 4°C with purified anti-OX40 mAb, washed, and reacted for 15 min at 4°C with anti-rat Ig microbeads (Miltenyi Biotec). Cells were washed, resuspended in PBS containing 2 mM EDTA and 0.5% BSA, and added to the magnetic column. The nonattached OX40− fraction was collected by washing the column with wash buffer; OX40+ cells were collected after detachment of the magnetic column. OX40+ and OX40− effector cells were adjusted to the same concentration and assayed for lytic activity in a 4-h redirected cytotoxicity assay.

For analyses of IL-2, IL-4, IL-10, and IFN-γ cytokine activity, 1 × 106 IELs were cultured for 24 h as described above in rIL-15-supplemented medium (rIL-2 was omitted to permit measurement of IL-2 activity). Cell-free supernatants were collected and assayed for cytokine activity with a commercial immunoassay kits (eBiosience, San Diego, CA) using the manufacturer’s protocols and standards.

Results

Murine CD8 IELs bear properties of partial T cell activation

Freshly isolated IELs were stained for CD69, OX40, and intracellular IFN-γ expression and were assayed for evidence of activation in the redirected cytotoxic assay. As shown in Fig. 1⇓, >98% of CD8+ IELs were CD69+; in the presence but not in absence of anti-CD3 mAb, those cells were cytotoxic for FcR-bearing target cells. However, interestingly, despite that evidence of activation, CD8+ IELs were OX40− and OX40L− and did not stain for intracellular IFN-γ (Fig. 1⇓). These were consistent findings from BALB/c and C57BL/6 mice. To understand how OX40 expression relates to the process of T cell activation, lymph node cells were activated in vitro by CD3 stimulation. These studies demonstrated that CD69 and OX40 expression are acquired at different times during activation. Note that between 24 and 48 h only 16–17% of CD8+ lymph node T cells expressed OX40, whereas 57–61% were CD69+, and that OX40 was not maximally expressed until 72 h poststimulation (Fig. 2⇓), a finding also reported by others for OX40 (10). These findings thus indicate that during immune activation it is possible to have a population of CD8 T cells that are CD69+ but OX40−, similar to freshly isolated IELs (Fig. 1⇓).

FIGURE 1.
FIGURE 1.

Freshly isolated CD8α IELs express CD69 and are cytolytic in redirected cytotoxicity assays despite the lack of expression of OX40, OX40L, and intracellular IFN-γ. Data are representative of three to eight experiments.

FIGURE 2.
FIGURE 2.

CD69 and OX40 are acquired at different times during CD3-mediated activation of peripheral T cells. Lymph node cells were stimulated in vitro with immobilized anti-CD3 Ab. Cells were collected at daily intervals or before culture and stained for expression of CD69, OX40, and CD8α. Data are representative of four experiments.

CD3 stimulation induces OX40 and OX40L expression on CD8+ IELs

We sought to determine the conditions that favor terminal activation of IELs. Freshly isolated IELs were cultured in vitro with or without CD3 stimulation in the presence of IL-2 and IL-15. IL-15 is known to promote the survival of IELs, particularly though not exclusively for CD8αα cells, whereas IL-2 is required as a growth factor for IELs following TCR/CD3 stimulation (21, 22). Optimal concentrations of rIL-15 for these experiments were determined by measuring the viability of non-CD3-stimulated IELs after 24 h of culture. Similar to what has been reported by others (21), 50–200 ng/ml of rIL-15 significantly enhanced IEL viability compared with cultures without IL-15 (data not shown). Viability was also determined for IELs at 24 and 48 h in the presence of 4 ng/ml rIL-2, a concentration previously used for in vitro IEL culture experiments (21), with and without rIL-15. Viability of IELs remained high (85–90%) when cultured in the presence of rIL-15. The presence of rIL-2 in those cultures did not alter viability, though viability was reduced slightly in cultures containing rIL-2 alone (data not shown). Concentrations of 4 ng/ml rIL-2 and 100 ng/ml rIL-15 were therefore used in subsequent experiments.

Expression of OX40 and OX40L on IELs was evaluated after 24 and 48 h of culture with rIL-2 and rIL-15 in the presence or absence of CD3 stimulation. Non-CD3-stimulated IELs (or IELs cultured with isotype-matched hamster mAb) cultured for 24 h retained levels of OX40 expression (Fig. 3⇓A) that were only slightly higher than those of freshly isolated cells (Fig. 1⇑). Similar findings were observed for OX40L expression in non-CD3-stimulated cultures (data not shown). In contrast, a significant proportion of CD3-stimulated IELs expressed OX40 and OX40L at 24 h; this consisted of a population of CD8+OX40high cells and CD8+OX40Lhigh cells, and a population of CD8+OX40low cells and CD8+OX40Llow cells that appear to be in the process of acquiring OX40 and OX40L expression (Fig. 3⇓, C and E). The patterns observed at 24 h were retained and extended in 48-h cultures in that nearly all CD8+ cells cultured with rIL-2 and rIL-15 without CD3 stimulation were OX40− and OX40L− (Fig. 3⇓B), whereas 98% of CD8+ cells acquired OX40 expression and 89% of the CD8+ cells acquired OX40L in 48-h CD3-stimulated cultures (Fig. 3⇓, D and F).

FIGURE 3.
FIGURE 3.

OX40 (C and D) and OX40L (E and F) expression is acquired on cultured IELs following stimulation with immobilized CD3, but is not expressed in the absence of stimulation (A and B). Cells were cultured in medium supplemented with 4 ng/ml rIL-2 and 100 ng/ml rIL-15. Data are representative of two to six experiments.

To confirm that OX40 and OX40L expression occurred on IELs and not epithelial cells, 24- and 48-h CD3-stimulated cells were stained for OX40 and OX40L expression with Ab to the LCA. As shown in Fig. 4⇓A, nearly all OX40+ and OX40L+ cells were LCA+. Moreover, OX40 and OX40L expression occurred on both CD8α and CD8β cells (Fig. 4⇓B).

FIGURE 4.
FIGURE 4.

A, Coexpression of OX40 and OX40L with the LCA on CD3-activated IELs confirms the expression of those markers on IELs. B, OX40 and OX40L expression is expressed on both CD8α+ and CD8β+ IELs. Data are representative of two experiments.

Unlike IELs, CD3-stimulated peripheral T cells do not express OX40L

The above findings suggest that OX40-OX40L communication in the gut epithelium takes place between CD8 T cells during activation. To determine whether OX40L expression is a feature of T cell activation that is unique to the gut, and whether the activation conditions used for IELs might have affected the expression of those markers, lymph node T cells were cultured under identical conditions of CD3 stimulation and rIL-2 and rIL-15 culture used for IELs. At both 24 (Fig. 5⇓, A and C) and 48 (Fig. 5⇓, E and G) h there was no significant increase in OX40L expression on lymph node T cells following CD3 stimulation. T cells cultured with rIL-1 and rIL-15 in the absence of CD3 stimulation also failed to express OX40L at 24 (Fig. 5⇓, B and D) and 48 (Fig. 5⇓, F and H) h. The lack of OX40L expression by activated lymph node T cells is consistent with studies by others (23), and they imply that OX40L expression on IELs is part of a specialized communication network in the intestinal epithelium.

FIGURE 5.
FIGURE 5.

Unlike IELs, lymph node T cells stimulated for 24 (A and C) or 48 (E and G) h with immobilized anti-CD3 in the presence of 4 ng/ml rIL-2 and 100 ng/ml rIL-15 do not express OX40L. Expression of OX40L on lymph node T cells cultured for 24 (B and D) or 48 (F and H) h with rIL-2 and rIL-15 in the absence of CD3 stimulation. Data are representative of three experiments.

OX40 expression on IELs is linked to IFN-γ synthesis, enhanced cell-mediated cytotoxicity, and suppressed IL-10 production

The combined expression of OX40 and OX40L on activated IELs suggests that an OX40-triggered signal is involved in the effector response of CD3-stimulated IELs. To determine how OX40/OX40L expression is functionally associated with IEL activation, three-color staining was done for OX40 and CD8 expression in conjunction with intracellular IFN-γ staining after overnight in vitro CD3-induced stimulation. Similar to freshly isolated IELs (Fig. 1⇑), IELs cultured for 16 h with rIL-2 and rIL-15 but without CD3 stimulation remained negative for intracellular IFN-γ staining (Fig. 6⇓). In contrast to unstimulated cells, approximately half of all CD8+ IELs from 16-h CD3-stimulated cultures were intracellular IFN-γ+. However, most important was the finding that OX40 expression was exclusively associated with the intracellular IFN-γ+ population of CD8 IELs (Fig. 6⇓), indicating that OX40 expression demarcates a population of activated CD8+ IELs with newly acquired IFN-γ synthesis.

FIGURE 6.
FIGURE 6.

Activation of IEL by CD3 stimulation leads to intracellular IFN-γ synthesis and OX40 expression on CD8α+ IELs. Analyses of cells were done from cultures stimulated with immobilized anti-CD3 mAb or unstimulated cultures after 16 h. Data are representative of two experiments.

Three experiments were done to determine how OX40 is involved in IEL effector activity. In the first, the influence of OX40 expression on the cell-mediated cytotoxic activity of IELs was studied. IELs were collected from 24-h CD3-stimulated cultures and were separated by MACS sorting into OX40+ and OX40− populations as described in Materials and Methods. Each group was assayed for lytic activity in redirected cytotoxicity assays. Twenty-four-hour cultures were used because some, though not all, CD8+ IELs are OX40+ cells and it was possible to obtain OX40+ and OX40− cells from the same culture preparation. Shown in Fig. 7⇓A, OX40− cells retained lytic activity that was very similar to that of freshly isolated cells; note that at an E:T ratio of 25:1, both freshly isolated (Fig. 1⇑) and OX40− IELs from CD3-stimulated cultures (Fig. 7⇓A) had 10–15% lytic activity in redirected assays. In contrast, OX40+ IELs following CD3 stimulation had consistently higher levels of lytic activity (67% in this experiment at an E:T ratio of 25:1) (Fig. 7⇓A).

FIGURE 7.
FIGURE 7.

A, Redirected cytotoxicity of OX40+ and OX40− IELs after 24-h culture with anti-CD3 mAb, demonstrating that OX40 expression correlates with enhanced cell-mediated cytotoxicity by IELs. Data are representative of two experiments. IFN-γ (B) and IL-10 (C) production by CD3-stimulated IELs cocultured with anti-OX40 or rat control Ig mAbs. Mean values ± SEM of data from three experiments. ∗, p < 0.025 comparing anti-OX40-treated and control-treated cells as determined by Student’s t test for unpaired observations. D, Increase in 24-h IL-10 production by CD3-stimulated IELs cultured in the presence of soluble anti-OX40L mAb (10 μg/ml) compared with CD3-stimulated IELs cultured with isotype-matched control mAb; mean value ± SEM of three experiments. E, Blockade effect of soluble anti-OX40L mAb (2–16 μg/ml) on 24-h IL-10 production from CD3-stimulated IELs compared with that of IELs stimulated with anti-CD3 plus control mAb, resulting in a dose-dependent increase in IL-10 synthesis with maximal effect at 8 μg/ml. Data are from one titration experiment. Note also the suppressed production of IL-10 by IELs costimulated with anti-OX40 mAb in conjunction with CD3 stimulation. Dashed lines indicate position of control Ig costimulation IL-10 level.

In the second set of experiments, cytokine activity was measured for four T cell cytokines (IL-2, IL-4, IL-10, and IFN-γ) from IEL cultures after 24-h stimulation in wells containing immobilized anti-CD3 and anti-OX40 mAbs or immobilized anti-CD3 and control Ig mAbs. No appreciable difference was observed for IFN-γ levels (Fig. 7⇑B) or for IL-2 or IL-4 (data not shown) for anti-OX40-costimulated vs control Ig cultures. These findings indicate that even though OX40+ cells are the principal source of IFN-γ (Fig. 6⇑), direct costimulation via OX40 does not affect the level of IFN-γ produced by those cells. However, of particular interest was the finding that anti-OX40 costimulation of CD3-stimulated IELs significantly reduced the level of secreted IL-10 (Fig. 7⇑C). Note that because more cells are OX40+ than are OX40L+ at 24 h (Fig. 3⇑), the experimental system used in this study detects the effects of OX40 ligation as soon as OX40 expression has occurred, thus accounting for the difference between anti-OX40-stimulated and control-stimulated cultures.

Finally, experiments were done to determine whether blockade of OX40 stimulation would alter IL-10 production. IELs were cultured with immobilized anti-CD3 and control mAbs, with immobilized anti-CD3 and anti-OX40 mAbs, or with immobilized anti-CD3 and soluble anti-OX40L mAbs. As shown in Fig. 7⇑D, blockade of OX40-OX40L binding using soluble anti-OX40L mAb (10 μg/ml) resulted in an increase of 22.4% in IL-10 production. These findings were confirmed using a range of soluble anti-OX40L mAb in CD3-stimulated cultures to block OX40-mediated activation; note the dose-dependent effect of anti-OX40L blockade with maximal effect (269.2% increase above anti-CD3 simulation alone) occurring at 8 μg/ml anti-OX40L mAb (Fig. 7⇑E). As before, anti-OX40 costimulation resulted in suppression of IL-10 synthesis (Fig. 7⇑E). These findings indicate that an OX40-triggered signal during CD3 activation of murine IELs leads to suppressed production of IL-10, and that the absence of OX40-OX40L interactions results in markedly elevated levels of IL-10 synthesis by IELs.

Discussion

What do these observations tell us about the process of IEL activation? The finding that OX40 is not present on IELs until after CD3 stimulation has taken place despite CD69 expression and cytolytic activity by those cells suggests that under normal conditions murine IELs are maintained in a remarkably dynamic yet controlled state of activation. Although it is possible that at any given time all IELs are merely in the process of proceeding toward a state of terminal activation, this seems unlikely because, if true, some cells should bear properties of fully activated T cells. Rather, our findings favor the likelihood that most IELs exist in a state of partial differentiation as seen by their potential for cytolytic activity even though they lack other activation-related properties, such as intracellular IFN-γ synthesis (Fig. 1⇑). The induction of intracellular IFN-γ activity upon CD3 stimulation is consistent with prior observations for IELs (24) and now links that functional response of IELs to a phenotypically defined effector population. Because IFN-γ is known for its wide range of immunobiological activities, indiscriminate IFN-γ production in the absence of antigenic threat locally would not be needed and in fact could be detrimental (25). However, rapid synthesis of IFN-γ upon recognition and destruction of Ag-bearing target cells would have a number of beneficial effects, including increases in MHC expression (26), local antiviral activity (27), and enhanced NK cell activity (28). Moreover, in this system immune destruction of Ag-bearing tissues would be achieved far more quickly than if activation were to take place from a completely resting state, as occurs for most extraintestinal peripheral T cells.

The presence of OX40L on CD8+ IELs after CD3 stimulation implies that the primary pathway of interactions for OX40 signaling to IELs occurs between CD8 T cells after TCR/CD3 stimulation has taken place. Although regulated OX40L expression is a feature of the DC response upon immune challenge (15), the potential for this as a means of OX40 signaling in the intestinal epithelium is unclear given that information regarding the distribution and numbers of DCs in the intestinal epithelium is scant. In fact, best evidence suggests that DCs are more sparsely located in the intestinal epithelium than in the lamina propria or Peyer’s patches (29). However, in that context, a process of intracellular molecular transfer of OX40L onto T cells has recently been described, involving the acquisition of OX40L on both CD4+ and CD8+ cells (30). Thus, it is possible that OX40L expression on IELs is not due to de novo synthesis by CD8+ IELs, but that it is acquired by the active shedding of OX40L from a minor component of OX40L+ cells within the total IEL population, possibly a subset of the gut DCs alluded to above. Regardless, our finding that both OX40 and OX40L are acquired on a significant proportion of IELs within 24 h of CD3 stimulation, resulting in greater efficiency in cytotoxic activity with concomitant IFN-γ synthesis, defines a system of direct costimulatory activation between resident CD8+ IELs.

A mechanistic model for the findings described in this work is shown in Fig. 8⇓. According to that model, IEL activation after TCR engagement and CD3 signaling leads to rapid up-regulation of OX40 and OX40L. Because under normal conditions this would require an Ag-stimulated event, changes in OX40 and OX40L expression would occur only on Ag-specific cells. CD3 signaling would lead to rapid IFN-γ production by OX40+ cells with a concomitant increase in cytotoxic activity. Additionally, OX40-OX40L binding would suppress the overall production of IL-10 by IELs, thereby further promoting cell-mediated immunity and enhancing Th1 responses (31). Because IL-10 is known for its counterregulatory effects on activities mediated by inflammatory cytokines and chemokines (32, 33, 34, 35) and reduces resistance to a wide range of intracellular pathogens (36, 37, 38), suppressed levels of IL-10 would favor the generation of local immunity during the early stages of local infection. However, sustained OX40/OX40L expression on IELs would adversely promote local inflammation and immunopathology by suppressing IL-10 synthesis (39).

FIGURE 8.
FIGURE 8.

Model for role of OX40 and OX40L in IEL activation. A TCR/CD3 (yellow) signal induces rapid up-regulation of expression of OX40 (green) and OX40L (red) on IELs, leading to synthesis and release of IFN-γ. Ligation of OX40 by OX40L expressed on activated IELs enhances Th1 responses and cell-mediated immunity and suppresses IL-10 production by IELs.

A remaining question regarding murine IELs pertains to the types of signals or events that initiate the initial phase of activation before Ag recognition by those cells. Candidates for this include bacterial LPS, endotoxin, and superantigens, all of which are found to one degree or another in the intestinal tract. However, after the initial activation steps have taken place, failsafe mechanisms would be needed to prevent widespread terminal activation of IELs. One possibility is that IELs are held in check by other differentiation/regulatory factors associated with mucosal tissues that have inflammation-associated immune-modulating activities such as IL-12 and TGFβ (40). Yet a system of immunity such as this, while perhaps logistically effective for dealing efficiently with the vast number of Ags that confront the intestinal mucosa, has obvious potential disadvantages in situations of Ag persistence or immune dysregulation that lead to widespread inflammation. The use of OX40 as a marker for discriminating IELs that have undergone recent activation should be valuable for precisely identifying effector populations involved in local immunity and for understanding changes in IEL populations associated with inflammation and immunopathology.

Footnotes

  • ↵1 This work was supported by National Institutes of Health Grant DK35566.

  • ↵2 Address correspondence and reprint requests to Dr. John R. Klein, Department of Basic Sciences, Dental Branch, University of Texas Health Science Center, Rm 4.133, 6516 M.D. Anderson Boulevard, Houston, TX 77030. E-mail address: John.R.Klein{at}uth.tmc.edu

  • ↵3 Abbreviations used in this paper: IEL, intraepithelial lymphocyte; LCA, leukocyte common Ag; DC, dendritic cell; OX40L, OX40 ligand.

  • Received March 29, 2001.
  • Accepted September 24, 2001.

References

  1. ↵
    Sillet, H. K., J. Southgate, P. D. Howdle, L. K. Trejdosiewicz. 1999. Expression of activation and costimulatory elements by human intestinal intraepithelial lymphocytes. Scand. J. Immunol. 50: 52
    OpenUrlCrossRefPubMed
  2. ↵
    Goodman, T., L. Lefrancois. 1989. Intraepithelial lymphocytes: anatomical site, not T cell receptor form, dictates phenotype and function. J. Exp. Med. 170: 1569
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Mosley, R. L., D. Styre, J. R. Klein. 1990. Differentiation and functional maturation of bone marrow-derived intestinal epithelial T cells expressing membrane T cell receptor in athymic radiation chimeras. J. Immunol. 145: 1369
    OpenUrlAbstract
  4. ↵
    Viney, J. L., P. Kilshaw, T. T. MacDonald. 1990. Cytolytic αβ/+ and γ/δ+ T cells in murine intestinal epithelium. Eur. J. Immunol. 20: 1623
    OpenUrlCrossRefPubMed
  5. ↵
    Lake, J. P., C. W. Pierce, J. D. Kennedy. 1991. CD8+ αβ or γδ T cell receptor-bearing T cells from athymic nude mice are cytolytically active in vivo. J. Immunol. 147: 1121
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Lefrancois, L., T. Goodman. 1989. In vivo modulation of cytolytic activity and Thy-1 expression in TCR-γδ+ intraepithelial lymphocytes. Science 243: 1716
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Paterson, D. J., W. A. Jefferies, J. R. Green, M. R. Brandon, P. Corthesy, P. Puklavec, A. F. Williams. 1987. Antigens of activated rat T lymphocytes including a molecule of 50,000 Mr detected only on CD4 positive T blasts. Mol. Immunol. 24: 1281
    OpenUrlCrossRefPubMed
  8. ↵
    Al-Shamkhani, A., M. L. Birkland, M. Puklavec, M. H. Brown, W. James, A. N. Barclay. 1996. OX40 is differentially expressed on activated rat and mouse T cells and is the sole receptor for the OX40 ligand. Eur. J. Immunol. 26: 1695
    OpenUrlCrossRefPubMed
  9. ↵
    Pippig, S. D., C. Pena-Rossi, J. Long, W. R. Godfrey, D. J. Fowell, S. L. Reiner, M. L. Birkeland, R. M. Locksley, A. N. Barclay, N. Killeen. 1999. Robust B cell immunity but impaired T cell proliferation in the absence of CD134 (OX40). J. Immunol. 163: 6520
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Akiba, H., H. Oshima, K. Takeda, M. Atsuda, H. Nakano, A. Nakajima, C. Nohara, H. Yagita, K. Okumura. 1999. CD28-independent costimulation of T cells by OX40 ligand and CD70 on activated B cells. J. Immunol. 162: 7058
    OpenUrlAbstract/FREE Full Text
  11. ↵
    Ohshima, Y., Y. Tanaka, H. Tozawa, Y. Takahashi, C. Maliszewski, G. Delespesse. 1997. Expression and function of OX40 ligand on human dendritic cells. J. Immunol. 159: 3838
    OpenUrlAbstract
  12. ↵
    Lane, P.. 2000. Role of OX40 signals in coordinating CD4 T cell selection, migration, and cytokine differentiation in T helper (Th)1 and Th2 cells. J. Exp. Med. 191: 201
    OpenUrlFREE Full Text
  13. ↵
    Weinberg, A. D., A. T. Vella, M. Croft. 1998. OX-40: life beyond the effector T cell stage. Semin. Immunol. 10: 471
    OpenUrlCrossRefPubMed
  14. ↵
    Brocker, T., A. Gulbranson-Judge, S. Flynn, M. Riedinger, C. Raykundalia, P. Lane. 1999. CD4 T cell traffic control: in vivo evidence that ligation of OX40 on CD4 T cells by OX 40-ligand expressed on dendritic cells leads to the accumulation of CD4 T cells in B follicles. Eur. J. Immunol. 29: 1610
    OpenUrlCrossRefPubMed
  15. ↵
    Murata, K., N. Ishii, H. Takano, S. Miura, L. C. Ndhlovu, M. Nose, T. Noda, K. Sugamura. 2000. Impairment of antigen-presenting cell function in mice lacking expression of OX40 ligand. J. Exp. Med. 191: 365
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Chen, A. I., A. J. McAdam, J. E. Buhlmann, S. Scott, M. L. Lupher, E. A. Greenfield, P. R. Baum, W. C. Fanslow, D. M. Calderhead, G. J. Freeman, A. H. Sharpe. 1999. Ox40-ligand has a critical costimulatory role in dendritic cell:T cell interactions. Immunity 11: 689
    OpenUrlCrossRefPubMed
  17. ↵
    Gramaglia, I., A. D. Weinberg, M. Lemon, M. Croft. 1998. Ox-40 ligand: a potent costimulatory molecule for sustaining primary CD4 T cell responses. J. Immunol. 161: 6510
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Souza, H. S., C. C. S. Elia, J. Spencer, T. T. MacDonald. 1999. Expression of lymphocyte-endothelial receptor-ligand pairs, α4β7/MAdCAM-1 and OX40/OX40 ligand in the colon and jejunum of patients with inflammatory bowel disease. Gut 45: 856
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Stuber, E., J. A., A. Buschenfeld, T. von Freier, T. Arendt, U. R. Folsch. 2000. The expression of OX40 in immunologically mediated diseases of gastrointestinal tract (celiac disease, Crohn’s disease, ulcerative colitis). Eur. J. Clin. Invest. 30: 594
    OpenUrlCrossRefPubMed
  20. ↵
    Stuber, E., A. Von Frier, D. Marinescu, U. R. Folsch. 1998. Involvement of OX40-OX40L interactions in the intestinal manifestation of the murine acute graft-versus-host disease. Gastroenterology 115: 1205
    OpenUrlCrossRefPubMed
  21. ↵
    Chu, D. L., S. S. Chen, T. S. Wu, S. C. Kuo, N. S. Liao. 1999. Differential effects of IL-2 and IL-15 on the death and survival of activated TCR-γδ+ intestinal intraepithelial lymphocytes. J. Immunol. 162: 1896
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Lai, Y. G., V. Gelfanov, V. Gelfanova, L. Kulik, S. L. Chu, S. W. Jeng, N. S. Liao. 1999. IL-15 promotes survival but not effector function differentiation of CD8+TCRαβ+ intestinal intraepithelial lymphocytes. J. Immunol. 163: 5843
    OpenUrlAbstract/FREE Full Text
  23. ↵
    Satake, Y., H. Akiba, K. Takeda, M. Atsuta, H. Yagita, K. Okumura. 2000. Characterization of rat OX40 ligand by monoclonal antibody. Biochem. Biophys. Res. Commun. 270: 1041
    OpenUrlCrossRefPubMed
  24. ↵
    Beagley, K. W., K. Fujihashi, A. S. Lagoo, S. Lagoo-Deenadaylan, C. A. Black, A. M. Murray, A. T. Sharmanov, M. Yamamoto, J. R. McGhee, C. O. Elson, H. Kiyono. 1995. Differences in intraepithelial lymphocyte T cell subsets isolated from murine small versus large intestine. J. Immunol. 154: 5611
    OpenUrlAbstract
  25. ↵
    Simpson, S. J., G. A. Hollander, E. Mizoguichi, D. Allen, A. K. Bhan, B. Wang, C. Terhorst. 1997. Expression of pro-inflammatory cytokines by TCRαβ+ and TCRγδ+ T cells in an experimental model of colitis. Eur. J. Immunol. 27: 17
    OpenUrlCrossRefPubMed
  26. ↵
    Steimle, J. C., C. A. Siegrist, A. Mottet, B. Lisowska-Grospierre, B. Mach. 1994. Regulation of MHC class II expression by interferon-γ mediated by the transactivator gene CIITA. Science 265: 106
    OpenUrlAbstract/FREE Full Text
  27. ↵
    Joklik, W. K.. 1990. Interferons. B. Fields, and D. M. Knipe, eds. Virology 383 Raven Press, New York.
  28. ↵
    Matera, L., C. M. Contarini, G. Bellone, B. Forno, A. Biglino. 1999. Up-regulation of interferon-γ mediates the enhancement of spontaneous cytotoxicity in prolactin-activated natural killer cells. Immunology 98: 386
    OpenUrlCrossRefPubMed
  29. ↵
    MacPherson, G. G., and L. Liu. Intestinal dendritic cells. In Dendritic Cells. M. T. Lotze and A. W. Thomson, eds. Academic Press, San Diego, p. 141.
  30. ↵
    Baba, E., Y. Takahashi, J. Lichtenfeld, R. Tanaka, A. Yoshida, K. Sugamura, N. Yamamoto, Y. Tanaka. 2001. Functional CD4 T cells after intercellular molecular transfer of OX40 ligand. J. Immunol. 167: 875
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Ito, S., P. Ansari, M. Sakatsume, H. Dickensheets, N. Vazques, R. P. Donnelley, A. C. Larner, D. S. Finbloom. 1999. Interleukin-10 inhibits expression of both interferon α- and interferon γ-induced genes by suppressing tyrosine phosphorylation of STAT1. Blood 93: 1456
    OpenUrlAbstract/FREE Full Text
  32. ↵
    de Waal Malefyt, R., J. Abrams, B. Bennett, C. Figdor, J. de Vries. 1991. IL-10 inhibits cytokine synthesis by human monocytes: an autoregulatory role for IL-10 produced by monocytes. J. Exp. Med. 174: 1209
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Fiorentino, D. F., A. Zlotnik, R. R. Mosmann, Howard M. H., A. O’Garra. 1991. IL-10 inhibits cytokine production by activated macrophages. J. Immunol. 147: 3815
    OpenUrlAbstract
  34. ↵
    D’Andrea, A., A. M. Aste, N. M. Valiante, X. Ma, M. Kubin, G. Trinchieri. 1993. Interleukin 10 (IL-10) inhibits human lymphocyte interferon-γ production by suppressing natural killer-cell stimulatory factor/IL-12 synthesis in accessory cells. J. Exp. Med. 178: 1041
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Marfaing-Koda, A., M. Maravic, M. Humbert, P. Galanaud, D. Emile. 1996. Contrasting effects of IL-4, IL-10, and corticosteroids on RANTES production by human monocytes. Int. Immunol. 8: 1587
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Wagner, R. D., N. M. Maroushek, J. F. Brown, C. J. Czuprynski. 1994. Treatment with anti-interleukin-10 monoclonal antibody enhances early resistance to but impairs complete clearance of Listeria monocytogenes infection in mice. Infect. Immunol. 62: 2345
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Dai, W. J., G. Kohler, F. Brombacher. 1997. Both innate and acquired immunity to Listeria monocytogenes infection are increased in IL-10-deficient mice. J. Immunol. 158: 2259
    OpenUrlAbstract
  38. ↵
    Denis, M., E. Ghardirian. 1993. IL-10 neutralization augments mouse resistance to systemic Mycobacterium avium infection. J. Immunol. 151: 5425
    OpenUrlAbstract
  39. ↵
    Steidler, L., W. Hans, L. Schotte, S. Neirynck, F. Obermeier, W. Falk, W. Fiers, E. Remaut. 2000. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science 289: 1352
    OpenUrlAbstract/FREE Full Text
  40. ↵
    McGhee, J. R., H. Kiyono. 1999. The mucosal immune system. W. E. Paul, ed. Fundamental Immunology 4th Ed.934 Lippincott-Raven, New York.
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section