and Costimulatory Requirements Cytokine Production, Activation ...

The Differentiated State of Intestinal Lamina Propria CD41 T Cells Results in Altered Cytokine Production, Activation Threshold, and Costimulatory Requirements1 Stephen D. Hurst,2* Cristine J. Cooper,3* Stephanie M. Sitterding,* Jung-hee Choi,* Robin L. Jump,† Alan D. Levine,† and Terrence A. Barrett4* Intestinal lamina propria (LP) CD41 T cells are memory-like effector cells that proliferate at relatively low levels and require high levels of TCR signaling and costimulation for full activation in vitro. To study LP CD41 T cell functional potential we used DO11.10 TCR transgenic (Tg) mice specific for the class II MHC-restricted OVA323–339 peptide and nontransgenic BALB/c mice. Activation of LP Tg1 T cells with Ag using mucosal explants induced high levels of IL-2, IL-4, and IFN-g. Culturing isolated LP cells with IL-12 enhanced IFN-g production and down-regulated IL-4 and IL-2, whereas addition of IL-4 maintained IL-4 production without inhibiting IFN-g production. Systemic administration of relatively high dose (HD; 100 nM) OVA323–339 peptide induced similar levels of bromodeoxyuridine (BrdU) incorporation by LP and splenic Tg1 T cells in vivo, whereas low dose (LD; 4.5 nM) peptide injections induced 4-fold greater levels of BrdU incorporation for LP compared with splenic Tg1 T cells. Coadministration of CTLA-4Ig reduced BrdU incorporation for splenic cells by 70% with HD and LD stimulation, but had little effect on LP responses to HD stimulation. Results of in vivo studies were confirmed in nontransgenic BALB/c mice using HD (200 mg) and LD (10 mg) anti-CD3 mAb1/2 CTLA-4Ig. These results suggest that LP T cells are differentiated effector cells that respond at high levels when activated with relatively low levels of Ag- and B7-mediated costimulation in vivo. The reduced activation threshold of LP T cells may facilitate responses to low levels of Ag derived from mucosal pathogens. The Journal of Immunology, 1999, 163: 5937–5945. he primary function of CD41 T lymphocytes localized to the intestinal lamina propria (LP)5 is to participate in protective immune responses that maintain integrity of the mucosal barrier (1). Lamina propria CD41 T cells promote Th1like immune responses against intracellular pathogens and support Th2-like humoral responses against extracellular pathogens by providing help to mucosal B cells and plasmablasts (2). Most LP T cells express an activated helper-inducer phenotype. Flow cytometric analysis has shown that the majority of CD41 LP T cells are negative for expression of L-selectin (3, 4). Furthermore, between 66 –96% of LP lymphocytes express the CD45RO isoform of the

T

*Veterans Administration Lakeside Medical Research Center and Department of Medicine, Section of Gastroenterology, Northwestern University Medical School, Chicago, IL 60611; and †Department of Medicine, Case Western Reserve University School of Medicine, Cleveland, OH 44106 Received for publication January 25, 1999. Accepted for publication September 20, 1999. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by U.S. Public Health Service Grants R29DK47073-01A1 and 2RO1DK47073-06A, a Department of Veteran’s Affairs Merit Review Award, and a research grant from the Crohn’s and Colitis Foundation of America. 2 Current address: DNAX Research Institute of Molecular and Cellular Biology, Palo Alto, CA 94304. 3

C.J.C. is co-first author.

4

Address correspondence and reprint requests to Dr. Terrence A. Barrett, Northwestern University, Med/GI S208, 303 East Chicago Avenue, Chicago, IL 60611. E-mail address: [email protected] 5 Abbreviations used in this paper: LP, lamina propria; PP, Peyer’s patch; MLN, mesenteric lymph node; Tg, transgenic; BrdU, bromodeoxyuridine; FCM, flow cytometry; HD, high dose; LD, low dose; MBP, myelin basic protein; NCS, normal calf serum.

Copyright © 1999 by The American Association of Immunologists

CD45 molecule associated with previously activated, memory-like T cells (5– 8). However, LP T cells do not express other surface markers typical of resting memory cells. In fact, expression of the early activation markers CD25 (IL-2R), CD69, the 4F2 lymphocyte activation Ag, and transferrin receptors suggest that CD41 LP T cells are chronically stimulated (9). In vitro mitogenic activation of LP T cells with CD28 or CD2-mediated costimulation induces low proliferative responses with relatively high production of IFN-g, IL-4, and IL-5 (5, 9 –12). Taken together, these findings suggest that the activation state of LP CD41 Th cells alters their functional responses to local Ags. Repeated stimulation induces the transition of naive CD41 T helper precursors to mature, activated Th cells that mediate secondary immune responses (13–16). Ag-primed subsets of Th cells have been referred to as effector T cells by analogy to CD81 cytolytic T lymphocytes (13, 14, 17, 18). These Th cell subsets express an activated surface phenotype, secrete high titers of a wide range of cytokines (IL-2, IL-4, and IFN-g), and can mediate helper activity for B cells (19 –21). Earlier work has also demonstrated that previously activated CD41 T cells required lower levels of costimulation and Ag compared with naive subsets (22, 23). These studies showed that naive T cells failed to produce cytokine when activated with low concentrations of Ag or mitogen or with low costimulation-competent APC such as resting B cells. In contrast, previously activated effector populations produced higher levels of cytokine at low Ag concentrations and could use a wide variety of APC populations, including resting B cells. Together, these data suggest that effector T cells may be distinguished from naive T cells by the requirement for lower levels of activation to induce functional responses. Observations that Ag experience reduces the activation threshold of CD41 effector T cells have not been supported by studies of LP T cell activation. Results of in vivo studies 0022-1767/99/$02.00

5938 suggest that LP T cells require relatively high levels of TCR-mediated and costimulatory signaling to induce responses (5, 9, 10, 12, 24, 25). Thus, studies of LP CD41 T cell responses have not agreed with previous models of memory-like effector T cell populations generated from naive precursors. To study the functional properties of LP CD41 T cells in an Ag-specific model we used DO11.10 TCR transgenic (Tg) mice specific for the class II MHC-restricted OVA323–339 peptide (26). Previous studies from our laboratory suggested that Tg1 T cells in the intestinal LP of DO11.10 mice were activated by environmental Ag through endogenous TCRs (27). Activation of LP Tg1 T cells via endogenous TCRs led to the generation of a population of effector CD41 T cells that expressed surface and functional phenotypes typical of CD41 LP T cells in normal murine and human systems (1, 3, 4, 6 –12, 25, 28). However, expression of Ag-specific TCR by LP Tg1 T cells allowed us to examine the effects of antigenic stimulation on an activated population of LP T cells. The current results show that CD41 Tg1 LP T cells produce IL-4 and IFN-g at relatively high levels and are resistant to additional differentiation in vitro. These results are consistent with those of Saparov et al. (29), who detected relatively high numbers of Tg1 LP T cells in DO11.10 mice positive for IFN-g or IL-10 using singlecell in situ hybridization. Furthermore, our studies of T cell responses in vivo using BrdU incorporation as a marker of activation suggest that LP cells are more responsive to low levels of TCR signaling and B7-mediated costimulation than are naive CD41 splenic T cells. Collectively, these findings suggest that previous activation of CD41 LP Th cells lowers the activation threshold and increases both expansion as well as cytokine production induced by antigenic stimulation.

Materials and Methods Animals BALB/c mice were obtained from the National Cancer Institute (Frederick, MD). D011.10 Tg1 mice (a gift from Dr. Dennis Loh, Nippon Research Center, Kamakura, Japan) were bred in facilities at the Lakeside Veteran’s Administration Medical Sciences Building and maintained under specificpathogen-free conditions. Transgenic mice were mated to BALB/c breeders for more than six generations, and the progeny was screened for expression of the TCR transgene by flow cytometry (FCM) using the clonotypic mAb KJ1-26.1 (27). DO11.10 Tg mice were used between 5 and 8 wk of age.

Cell isolation Splenic cells were isolated as previously described (30). Briefly, spleens were mechanically disassociated and RBC lysed in ACK lysis buffer. Cell suspensions were washed and stored in DMEM with 10% FCS (10% DMEM) on ice until used. Cell preparations were purified by mAb and complement lysis as described below. Lamina propria cells were prepared by modifications of a protocol previously described (12, 27). Briefly, small intestines were removed and flushed with cold PBS to remove fecal contents. After PP were excised, intestines were opened longitudinally, minced into 5- to 10-mm pieces, and washed extensively with cold PBS. Mucosal pieces were then digested twice for 30 min each time at 37°C in trypsinizing flasks with 5 mM EDTA (Sigma, St. Louis, MO) and 10% normal calf serum (NCS) (Life Technologies, Grand Island, NY) in PBS. After each digestion with EDTA, mucosal pieces were washed with cold PBS, and the supernatants were discarded. The remaining tissue was then digested for four 30-min intervals in a buffer containing 100 U/ml collagenase (Sigma), 25 mM HEPES, 7 mM CaCl2, and 20% NCS in DMEM. After each 30-min interval, the cells released were centrifuged, washed, and stored in 10% DMEM on ice, and the mucosal pieces were replaced in the collagenase buffer. Following the fourth interval, the digestion buffer was supplemented with an additional 100 U/ml of collagenase and 10% NCS and digested for two additional 30-min intervals. In some cases, pieces were digested with 400 U/ml of collagenase (Roche Molecular Biochemicals, Indianapolis, IN). Large debris was eliminated from the cell suspension by passage through nylon wool columns at room temperature. Column elutant was diluted in 5% DMEM containing 0.3 mg/ml DTT (Life Technologies), and viable cells were isolated by centrifugation over Nycoprep 1.077 (Ac-

FUNCTIONAL DIFFERENTIATION OF LP T CELLS curate Chemical, Westbury, NY). After centrifugation, cells were collected from the interface, washed, and pelleted. Cells were then purified further by mAb and complement lysis as described below. Splenic and LP Tg1 T cells were purified by mAb and complement lysis as previously described (31). Briefly, isolated cells were incubated on ice in a mixture of J11d.2 (anti-heat-stable Ag), 2.43 (anti-CD8), M5/114 (anticlass II), and 2.4G2 (anti-Fc receptor). Following a 30-min incubation with these mAbs, cells were washed and incubated on ice with MAR 18.5 (mouse anti-rat Ig; all from American Type Culture Collection, Manassas, VA) for 30 min. Cells were then washed and incubated with a 1/10 dilution of rabbit Low-Tox complement (Accurate Chemical) for 45 min at 37°C, and viable cells were recovered on Nycoprep 1.077 gradients. Residual mAb-labeled cells were eliminated by incubation with anti-rat Ab-coated magnetic beads (40 beads/1 T cells; Dynal, Lake Success, NY), followed by magnetic negative selection of the bound cells. The resulting T cell preparation was .90% pure for CD41 T cells by FCM (data not shown). These cells were used as purified Tg1 T cells.

Culture conditions T cells were cultured in DMEM (Life Technologies) supplemented with 5% FCS (Life Technologies), 2-ME, penicillin-streptomycin, HEPES, and L-glutamine as previously described (31). Activation experiments were performed as described with modification (31). Briefly, 5 3 104 purified Tg1 T cells were cultured with 2.5 3 105 irradiated, T cell-depleted, BALB/c splenic APC and OVA323–339 peptide (Bio-Synthesis, Lewisville TX) in flat-bottom 96-well plates. Unless otherwise indicated, OVA323–339 peptide was used at 300 pM for activation. Murine CTLA-4Ig (a gift from Dr. Peter Linsley) that had been mutated at the complement binding region was used at 10 mg/ml where indicated (32). To assess the effects of cytokine on the functional differentiation of Tg1 T cells, purified Tg1 splenic or LP T cells (1 3 106 cells/well) were activated with APC and Ag alone or with exogenous rIL-4 (Genzyme, Cambridge, MA) at 100 U/ml or rIL-12 (a gift from Dr. Frank Fitch) at 5 U/ml for the first round of stimulation only. In some cultures the rat isotype control (anti-b-galactosidase) or neutralizing anti-IFN-g mAb XMG-1.2 was added (10 ng/ml). The effects of cytokine added to initial cultures were determined by assessing cytokine levels 48 h after tertiary stimulation in the absence of exogenous cytokine.

In situ LP culture Excised small bowel from unimmunized Tg1 mice was flushed with cold PBS and opened longitudinally. Uniform tissue samples (28 mm) (26) were isolated using a sterile, 6-mm diameter boring tool (Baxter Healthcare Corp., McGaw Park, IL), washed with PBS, and incubated with EDTA and 10% NCS in PBS for 30 min. Tissue explants were then washed in medium and cultured with OVA323–339 peptide (300 pM).

Cell proliferation For proliferation studies, spleens were teased apart, digested with collagenase (400 U/ml; Roche Molecular Biochemicals) for 2 h and then filtered through a 100-mm pore size nylon mesh (Falcon, Franklin Lakes, NJ) to yield a single-cell suspension. Lymphocytes were isolated by density centrifugation over Histopaque-1077 (Sigma), collected from the interface, and washed before culture. Lamina propria lymphocytes were isolated by collagenase digestion as described above and enriched by density centrifugation, layering the single-cell suspension over 30% Percoll (Pharmacia Biotech, Piscataway, NJ). Splenic or LP mononuclear cells (1 3 105 cells/ well) were washed before culture in U-bottom 96-well tissue culture plates (Falcon) for a total of 72 h with stimuli added at culture initiation. To measure proliferation, [3H]thymidine (0.5 mCi/well) was added for the final 24 h of incubation. [3H]thymidine incorporation was measured in harvested cells using a scintillation counter (Wallac, Gaithersburg, MD).

Cytokine analysis To assess cytokine production for isolated cells, 5 3 104 T cells were cocultured in 24-well plates with 2.5 3 105 irradiated APC and Ag. Unless otherwise indicated, supernatants were collected at 48 h and were analyzed for IL-2, IL-4, and IFN-g. Supernatants were stored at 220°C until analysis by ELISA. The ELISAs for IL-2, IL-4, IL-10, and IFN-g were performed according to manufacturer’s specification (Endogen, Cambridge, MA). Wells were developed with ABTS soluble substrate (Zymed, South San Francisco, CA) and read at 415 nm on a spectrophotometer. Cytokine levels are presented as units per milliliter according to previously described conversion values where 1 U IL-2 5 14.2 pg, 1 U IL-4 5 0.7 pg, 1 U IL-10 5 2 pg, and 1 U IFN-g 5 100 pg (33). The limits of detection for these ELISAs were as follows: IL-2, 0.2 U; IL-4, 4 U; IL-10, 1 U; and IFN-g, 0.5 U.

The Journal of Immunology

5939 Lamina propria Tg1 T cells resist Th2 differentiation to exogenous rIL-4 by maintaining IFN-g but down-regulating IL-2 production

FIGURE 1. Mucosal explants produce cytokine in response to Ag in vitro. Mucosal tissue samples of equal size (28 mm2) (26) were incubated with EDTA to remove the epithelial cell layer as described in Materials and Methods. Explants were then cultured in medium alone or with OVA323–339 peptide (300 pM). Cytokine production was determined by ELISA at 48 h and is presented as units per milliliter. Data shown are representative of three experiments.

Bromodeoxyuridine incorporation analysis Bromodeoxyuridine analysis was achieved by variations of protocols first presented by Carayon and Bord (34) and Tough and Sprent (35). Mice were given 200-ml i.p. injections of sterile, endotoxin-free PBS diluent with control myelin basic protein (MBP), OVA323–339 peptide, control hamster IgG (PharMingen, San Diego, CA), or anti-CD3 (2C11, PharMingen) at the amounts indicated. In some experiments mice were given i.p. injections of 0.5 mg of murine control-Ig or CTLA-4Ig 12 h before i.p. OVA323–339 or anti-CD3. Twelve hours before sacrifice, mice were given 1 mg of BrdU (Sigma) in 200 ml of sterile PBS i.p. This 12 h point was used to limit the background and maximize the signal observed in previously published methodologies. At sacrifice, lymphocytes were prepared as previously described (27). Cells were stained with anti-CD4-PE (PharMingen) and KJl-26.1-BIO (antiTg TCR, a gift from Dr. Phillipa Marrack, National Jewish, Denver, CO). Biotinylated KJl-26.1 was visualized with strepavidin-CyChrome (PharMingen). Cells were then fixed overnight in 1% electron microscopy grade paraformaldehyde (Polysciences, Warrington, PA) in PBS with 0.2% Tween-20 (Sigma), washed with cold PBS, and resuspended in 500 ml of Mg-Ca DNase buffer (5 mM MgCl2 and 5 mM CaCl2 in 13 PBS). Fifty Kunitz units of DNase I (Roche Molecular Biochemicals) were added, and the samples were incubated for 1 h at 37°C. Samples were then washed with FACS buffer containing 0.2% Tween-20, and 15 ml of FITC-conjugated murine control (PharMingen) or anti-BrdU (Becton Dickinson, San Jose, CA) was added before incubation for 45 min at 4°C. Samples were washed with FACS buffer containing 0.2% Tween-20 and were analyzed by FCM. Events were collected and analyzed using a Becton Dickinson FACScan and LYSIS II software.

Results Stimulation of cytokine production in mucosal explants Results of previous studies suggested that IL-4 and IFN-g cytokine production had been induced in LP T cells as a consequence of activation in the intestinal microenvironment (27). However, in that study LP Tg1 T cell cytokine production was assessed using splenic APC and OVA323–339 peptide. To assess cytokine production in response to Ag presented locally by mucosal APC we used an in situ explant system. After removal of the epithelial layer with EDTA digestion, uniformly small intestinal tissue samples were cultured in 24-well plates with control or OVA323–339 peptide (300 pM) and assayed for cytokine production at 24 h. This culture system preserved the normal microenvironment of LP APC and Tg1 T cells. As shown in Fig. 1, incubation with Ag stimulated the production of high amounts of IL-2, IL-4, and IFN-g similar to levels we had previously observed using purified LP T cells (27), while cytokine production was not detected in control wells. These data suggest that cytokines produced by isolated LP Tg1 T cells with splenic APC and Ag reflect the cytokine profile produced by LP T cells in response to Ag and local APC.

Based on studies by Bucy et al. (36, 37) as well as our own intracellular cytokine staining (data not shown) we suspected that activated Tg1 T cells in the LP were a mixed population of Th1and Th2-like cells. Reports by Murphy and colleagues (38) and others have suggested that Th1 and Th2 cells exhibit distinct lineage commitment. In these studies the characteristic cytokine profile of Th2 cells was more stable than that of Th1 cells with repeated stimulation in the presence of differentiating cytokine in vitro. To address the lineage commitment of LP cells, freshly isolated splenic and LP T cells from DO11.10 mice were cultured with APC, OVA323–339 peptide, rIL-4, OVA323–339 peptide and rIL-12, or medium alone. After initial activation with OVA323–339 peptide and exogenous cytokine, the cells were stimulated twice without exogenous cytokine for a total of three stimulations in vitro. Following the tertiary stimulation, the cultures were activated with OVA323–339 peptide and splenic APC without exogenous cytokine to determine the cytokine production profile. Fig. 2 shows that, as we had previously described, splenic Tg1 T cells produced relatively high levels of IL-2 and low levels of IFN-g and IL-4 upon primary stimulation. Addition of rIL-12 to splenic Tg1 T cell cultures promoted a Th1 profile of cytokines with high levels of IL-2 and IFN-g without IL-4. In contrast, addition of exogenous rIL-4 induced a Th2 profile with high IL-4 and reduced IL-2 and IFN-g. Activation of LP Tg1 T cells cultured with rIL-12 up-regulated the production of IFN-g and down-regulated IL-4 production compared with the primary stimulation. However, very little IL-2 or IL-4 was produced from the rIL-12-stimulated LP T cell cultures, suggesting that IL-12 and IFN-g had effectively suppressed IL-4 production. Addition of IL-2 during initial cultures failed to reverse the trend toward diminished IL-2 production in later passages (data not shown), suggesting that consumption of IL-2 did not account for the extinction of IL-2 production detected in LP cultures. The down-regulation of IL-2 by rIL-12 observed for LP, but not for spleen, cells suggested that previous activation of LP T cells in vivo may have predisposed this population to down-regulate IL-2 synthesis. Culture with exogenous rIL-4 induced high IL-4 production by Tg1 LP cells, but a lower level than that of rIL-4- treated splenic Tg1 T cells. In contrast to splenic Tg1 T cells, LP Tg1 T cells continued to produce high levels of IFN-g upon addition of rIL-4. LP Tg1 T cell IFN-g production was, in fact, up-regulated from levels observed in primary cultures and suggested that rIL-4 had not effectively suppressed IFN-g production, as had occurred in splenic cultures. We also found that addition of neutralizing anti-IFN-g (XMG-1.2) mAb to initial cultures failed to reduce subsequent IFN-g production in IL-4-treated wells of LP T cells (data not shown). These data indicate that LP Tg1 T cells resist the effects of exogenous rIL-4 on IFN-g production. However, in response to rIL-12, Tg1 cells uniformly differentiate into a phenotype dominated by high IFN-g and low IL-2 and IL-4 production. These findings are consistent with data reported by Saparov et al. (29), who used single-label in situ hybridization of cytokine mRNA to show that Th1, but not Th2, LP T cells are resistant to phenotype reversal in vitro. In this report the authors also detected IL-10 and IL-5 production by LP cells, which was detected at low levels in our studies (data not shown). Differences in the detection of IL-10 and IL-5 may have been due to differences in culture conditions and/or sensitivities of the cytokine assays. Taken together, these data were consistent with the hypothesis that the induction of Th1 cytokines produced by LP Tg1 T cells is resistant to suppression by Th2-promoting cytokines and

5940

FUNCTIONAL DIFFERENTIATION OF LP T CELLS

FIGURE 2. Cytokine production from Tg1 T cells differentiated with exogenous rIL-12 or rIL-4. Splenic or LP Tg1 T cells were activated with splenic APC and OVA323–339 peptide in the presence of rIL-12 (5 U/ml) or rIL-4 (100 U/ml) and were cultured for 10 days before restimulation with splenic APC and OVA323–339 peptide without exogenous cytokine. Ten days following the tertiary activation, cells were washed and activated with fresh APC. Cytokine production was determined at 48 h, by ELISA, and presented as units per milliliter. Cytokine profiles of activated freshly isolated Tg1 T cells are shown for reference. Data shown are representative of three experiments.

may explain why some cytokine knockout animal models that favor emergence of Th1 cells lead to chronic inflammation in the intestine. Proliferative responses of Tg1 LP T cells to Ag and anti-CD3 mAb Previous publications had reported that LP T cells were hypoproliferative in response to activation in vitro (11) and needed high levels of stimulation via TCR and CD28 pathways for optimal responses (5). To compare T cell responses in DO11.10 mice to previously published data in nontransgenic systems, proliferative responses were measured for splenic and LP lymphocytes stimulated by Con A, anti-CD3 mAb, and OVA323–339 peptide1/2 antiCD28 mAb. The results presented in Fig. 3 indicate that DO11.10 LP cells proliferated at greatly reduced rates compared with splenocytes for all stimuli studied. Proliferative responses induced by OVA323–339 peptide and anti-CD3 mAb were similar in LP cultures, whereas splenic proliferative responses were greatest in cultures activated with OVA323–339 peptide. Although LP T cells proliferated at relatively low levels in all cultures, the provision of anti-CD28 mAb tended to enhance responses regardless of the stimulus. A more clear positive effect of anti-CD28 was seen in splenic cultures activated with anti-CD3 or OVA323–339 peptide. CD28 stimulation alone does not promote proliferation (data not shown). Thus, Tg1 LP T cells were hypoproliferative and dependent on anti-CD28 for optimal responses in vitro. These findings confirmed the idea that Tg1 LP cells in DO11.10 mice responded in a similar pattern as LP cells in nontransgenic mice as well as primate and human models (5, 10, 12, 39).

Decreased activation threshold for LP T cells T cells may encounter intestinal Ag in vivo at concentrations far below that which is optimal for complete activation. Since previous work suggested that activated CD41 T cells from peripheral lymphoid tissue respond to lower levels of Ag than naive T cells (17, 18, 23), we examined the differential sensitivities of LP and splenic T cells to a range of OVA323–339 peptide concentrations. At optimal levels of OVA323–339 (300 pM), IL-2 production was greater for splenic compared with LP T cells. However, at suboptimal Ag concentrations from 3–30 pM (Fig. 4), LP T cells produced greater levels of IL-2 compared with splenic T cells. In fact, at Ag levels 10-fold below the concentration used to maximally stimulate IL-2 production, LP Tg1 T cells produced from 5- to 20-fold the amount of IL-2 secreted by splenic Tg1 T cells. These results suggest that LP Tg1 T cells required a lower threshold of TCR-mediated activation to induce functional responses. To test the responses of LP cells to high and low levels of Ag in vivo, we assessed BrdU incorporation in mice given i.p. injections of a range of OVA323–339 peptide dosages. The percentages of BrdU1 Tg1 CD41 T cells detected in at least three separate experiments were analyzed at each dosage level and reported as the mean 6 SEM). As shown in Fig. 5, high doses of systemic OVA323–339 peptide induced BrdU incorporation in relatively high proportions of Tg1 T cells recovered from both LP and splenic compartments. However, at the four lowest doses of OVA323–339 peptide examined (,40 nmol), the proportions of BrdU1 Tg1 CD41 T cells were significantly higher for LP compared with splenic populations. At the lowest level at which splenic Tg1 T


5941

FIGURE 3. Effect of CD28 costimulation on proliferation of Tg1 lymphocytes. Splenic and LP mononuclear cells were activated with anti-CD3 Ab (0.5 mg/ml) or OVA323–339 peptide (0.5 mM) in the presence or the absence of a stimulatory anti-CD28 Ab (0.5 mg/ml). Conconavalin A (5 mg/ml) served as a positive control. Proliferation was measured using [3H]thymidine incorporation as described in Materials and Methods and is presented as the mean 6 SEM. Data are representative of four experiments.

cell responses were detected (4.5 nmol), the proportion of LP BrdU1 Tg1 CD41 T cells was more than twice that of splenic cells. These in vivo data parallel IL-2 production observed in vitro (Fig. 4) and suggest that compared with resting, naive splenic CD41 T cell populations, LP CD41 T cells require a lower activation threshold to induce cytokine production and proliferation.

ive B cells and epithelial cells (1, 11, 40 – 43). Previous studies, however, have reported that LP T cells are highly dependent on costimulation for activation (5, 24, 39). Since these systems used mitogen-induced, non-Ag-mediated systems for LP T cell activation, we assessed the role of B7-1/2-mediated costimulation in TCR:OVA Ag-mediated stimulation of LP Tg1 T cells.

Dependence of LP Tg1 T cell cytokine production and in vivo proliferation upon B7-1/2-mediated costimulation Lamina propria T cells may encounter Ag presented by professional APC such as resident macrophages and dendritic cells as well as nonprofessional, costimulation-deficient APC such as na-

FIGURE 4. Sensitivity of Tg1 T cells to low levels of Ag. Purified splenic and LP Tg1 T cells were activated with splenic APC and various concentrations of OVA323–339 peptide as indicated. IL-2 production at 24 h was determined by ELISA. Data shown are representative of three experiments.

FIGURE 5. Reduction in the activation threshold of LP T cells in vivo. In vivo proliferation (BrdU incorporation) of Tg1 T cells following administration of low amounts of Ag. Mice were given a range of OVA323–339 peptide in decreasing amounts in PBS by i.p. injection. Thirty-six hours after OVA323–339 peptide administration and 12 h before sacrifice, mice were pulsed with 1 mg of BrdU in PBS i.p.12 h before sacrifice. Cell samples were stained for CD4 and KJ1-26 and intracellular BrdU and analyzed by flow cytometry. Data presented are the percentages of BrdU1 Tg1 T cells within a CD41 gate. Each point represents the mean of at least three experiments 6 SEM. Control levels (dashed line) indicate the BrdU incorporation for Tg1 CD41 T cells in mice given control (MBP) peptide with SEM 6 2%. p, p , 0.05 for differences between LP and spleen cells tested with paired two-tailed Student’s t test.

5942

FUNCTIONAL DIFFERENTIATION OF LP T CELLS

FIGURE 6. Effect of CTLA-4Ig on cytokine production by Tg1 T cells. Splenic and LP Tg1 T cells were cultured with splenic APC and OVA323–339 peptide with murine control-Ig or with APC with peptide and CTLA-4Ig at 10 mg/ml. Supernatants were harvested and analyzed for cytokine at 48 h as described in Materials and Methods. The production of IL-2, IL-4, and IFN-g is shown for cultures of Tg1 T cells with APC, Ag, and CTLA-4Ig compared with APC and Ag for freshly isolated cells. Data shown are representative of four experiments. The SE in each was ,15%.

Cytokine levels were measured in splenic and LP cell cultures activated in the presence of murine control-Ig or CTLA-4Ig. CTLA-4Ig binds to B7-1 and B7- 2 and blocks CD28 and CTLA-4 ligation (32). The results in Fig. 6 indicate that CTLA-4Ig reduced IL-2 levels for both splenic and LP populations by .90% with minimal effect on IFN-g production. Addition of CTLA-4Ig reduced IL-4 production for splenic populations by .50%, but reduced LP IL-4 levels by about 5%. Thus, the production of both IL-4 and IFN-g effector cytokines by LP Tg1 T cells was not significantly affected by inhibition of B7-mediated costimulation. As T cell activation in vivo may be induced with suboptimal levels of Ag, we examined the effect of blocking costimulation in vivo using high dose (HD) and low dose (LD) OVA323–339 peptide in DO11.10 mice (Figs. 7A and 8A). CTLA-4Ig was used to assess the contribution of B7-1/B7-2 ligands to T cell responses to Ag stimulation in vivo. DO11.10 mice were administered either HD (100 nmol) or LD (4.5 nmol) of OVA323–339 peptide in PBS. Representative flow cytometric data shown in Fig. 8B indicate that the percentages of BrdU1 Tg1 T cells in mice given HD and LD control peptide (MBP) were relatively low for both LP and splenic cultures (1–3%). T cells were stimulated with or without CTLA4Ig (0.2 mg) followed by injections of BrdU (1 mg) to label responding T cells. The results presented in Figs. 7A and 8B indicate that in response to LD OVA323–339 peptide administration, CTLA4Ig reduced the proportion of BrdU1 Tg1 T cells in the spleen by 70% ( p , 0.05) and 55% ( p , 0.01) in the LP. Administration of CTLA-4Ig in combination with high dose OVA323–339 peptide also reduced the proportion of BrdU1 Tg1 T cells in the spleen by 66% ( p , 0.01) vs OVA323–339 peptide alone. However, the effect of CTLA-4Ig on high dose OVA323–339 peptide stimulation was negligible (49% compared with 52%; nonsignificant) in the LP. Taken together, these data indicate that LP Tg1 T cells respond optimally in the absence of costimulation when activated with relatively high levels of TCR-mediated signaling. However, at suboptimal levels of Ag, costimulation is needed for complete activation. Observation of in vivo T cell responses in DO11.10 mice suggested that LP cells responded better to LD stimulation than spleen and were less dependent on CD28-mediated costimulation with HD stimulation. To test whether these findings could be generalized to nontransgenic systems, BrdU incorporation was assessed in Thy-1 gated splenic and CD41 LP T cells in normal BALB/c mice. BALB/c mice were injected with LD (10 mg) or HD (200 mg)

FIGURE 7. Effect of CTLA-4Ig on T cell BrdU incorporation in vivo. Mice were administered 0.5 mg of murine control-Ig or CTLA-4Ig as indicated, followed 12 h later by administration of HD or LD T cell stimulation. DO11.10 (A) mice were given HD (100 nmol or 150 mg) or LD (4.5 nmol or 6.75 mg) OVA323–339 peptide in PBS i.p. For comparison, nontransgenic BALB/c (B) mice were given HD (200 mg) or LD (10 mg) anti-CD3 (2C11) mAb i.p. Thirty-six hours after OVA323–339 peptide or anti-CD3 administration and 12 h before sacrifice, mice were pulsed with 1 mg of BrdU in 200 ml of PBS i.p. and then analyzed. A, Samples were stained for CD4 and KJ1-26 and intracellular BrdU and analyzed by flow cytometry. B, Cells were stained for Thy-1 and CD4 and intracellular BrdU and analyzed by flow cytometry. p, p , 0.05 for CTLA-4Ig vs control-Ig treated cells tested with paired Student’s t test. Each data point represents at least three experiments and is represented as the mean 6 SEM.

anti-CD3 mAb (2C11) and sacrificed 48 h later (Figs. 7B and 8B). These data confirm results in DO11.10 mice and show that nontransgenic CD41 LP cells respond nearly twice as well to LD stimulation as splenic cells. As in DO11.10 mice, CTLA-4Ig treatments reduced splenic CD41 T cell responses to HD and LD stimulations, whereas no effect was detected on LP T cell responses activated with HD anti-CD3. CTLA-4Ig reduced LP T cell responses to LD stimulation by nearly 50%. These data suggest that the contribution of costimulation to LP CD41 T cell activation was greater at lower levels of TCR signaling. Interestingly, CTLA-4Ig had little effect on splenic and LP CD81 (Thy-11, CD42) T cell responses to HD stimulation. These data are consistent with previous observations (44) and suggest that CD81 T cells are less dependent on costimulation than naive CD41 T cells. Responses to control hamster mAb injections in BALB/c mice were similar to levels observed in DO11.10 mice given control (MBP) peptide (data not shown). Taken together these data confirm findings in


FIGURE 8. Flow cytometric analysis of splenic and LP T cell activation in vivo. Representative flow cytometric data are presented from the studies summarized in Fig. 7. Mice were given 0.5 mg of murine control-Ig or CTLA-4Ig as indicated, followed 12 h later by administration of HD or LD T cell stimulation. DO11.10 (A) mice were given HD or LD control (MBP) or OVA323–339 peptide i.p. and 1 mg of BrdU i.p. 12 h before sacrifice. Cells were stained for CD4, KJ1-26, and intracellular BrdU as described in Fig. 7. Results for CD4 gated populations are shown. BALB/c (B) mice were given HD (200 mg) or LD (10 mg) anti-CD3 (2C11) mAb i.p. before BrdU labeling. Isolated cells were stained for Thy-1, CD4, and BrdU. Results for Thy-1gated populations are shown. The percentage of positively stained cells, based on control staining, is shown in each quadrant. Data shown are representative of three experiments.

DO11.10 mice and suggest that LP T cells in nontransgenic systems require lower levels of activation and costimulation compared with naive splenic CD4 T cell populations.

Discussion These results support the view that prior activation of LP CD41 T cells enhances the efficiency of effector immune responses in the

5943 intestine. In previous studies we reported that increased cytokine production correlated with expression of an activated profile of surface markers for Tg1 LP T cells in DO11.10 mice (27). Enhanced production of both IL-4 and IFN-g by LP T cells compared with that by splenic T cells suggests that previous Ag experience in the intestine promotes both cellular as well as humoral effector responses to subsequent stimulations. Production of IFN-g enhances class I and II MHC expression, up-regulates inducible nitric oxide synthase, and up-regulates microbicidal activities of macrophages (42, 45– 47). These effects enhance responses to intracellular enteric pathogens. The dominance of IFN-g responses for LP T cells was evidenced by the production of 35-fold greater levels of IFN-g by LP compared with splenic cells (27). Given the high levels of IFN-g produced by LP T cells one might expect that IL-4 production would be suppressed. However, relatively high levels of IL-4 were detected for freshly isolated LP cells activated in cultures as well as in mucosal explants. In studies by Saparov et al. (29), single-cell in situ hybridization was used to show that relatively high numbers of IL-10-positive CD41 T cells were present in the LP. Thus, it is likely that mechanisms in the mucosa operate to support the coexistence of IFN-g as well as IL-4 and IL-10 production. One possibility, suggested by Rennick and colleagues, is that endotoxin up-regulates IFN-g and IL-12, whereas local APC as well as T cells produce IL-10 that impairs the effects of IFN-g and IL-12 on Th cell differentiation (48, 49). Without IL-10, as in IL-10 knockout mice, the unopposed induction of Th1 responses leads to autoaggressive intestinal inflammation. Thus, our data suggest that activation of CD41 T cells in the LP promotes the induction of helper cytokines that support effector immune responses as well as regulate the balance between Th1 and Th2 cytokines. The results of in vivo studies suggest that the threshold of Agand B7/CD28-mediated costimulation is reduced for CD41 LP T cells compared with that for naive splenic T cells. These findings parallel those of Swain and colleagues, who described the properties of an activated, memory-like Th cell population that made increased levels of cytokine, provided B cell help, and was less dependent on costimulation for activation compared to naive T cells (23, 50). Our in vitro results extend these findings by demonstrating that CTLA-4Ig had little effect on IL-4 and IFN-g levels (Fig. 6). Using in vivo methods of assessing T cell responses, we observed that LP Tg1 T cell responses were independent of B71/B7-2-mediated costimulation at HD Ag stimulation. Thus, at high levels of TCR-mediated signaling, LP T cells may not require a second signal delivered via CD28. However, whether other costimulatory molecules play a role is unclear. At relatively low levels of TCR signaling, LP T cells require the contribution of CD28mediated costimulation to surpass the threshold needed for activation. At this point it is difficult to determine whether these results were due to a direct effect of Ag experience on LP T cells or to Ag-nonspecific mechanisms (e.g., via adhesion molecules and/or local cytokines; discussed below). The reduction in the level of Ag and costimulation needed to activate LP T cells may allow this population to be more sensitive to low levels of bacteria presented by mucosal APC such as epithelial cells, which lack B7 expression (40). The results of robust proliferative responses for LP T cells measured in vivo challenge current paradigms of mucosal T cell activation. Our results suggest that LP T cells respond as well or better than splenic T cells to in vivo antigenic stimulation. These results contrast observations of hypoproliferative responses detected in vitro for human (5, 10, 16) and murine LP T cells (Fig. 3). Low rates of proliferation reported in previous studies may be related to the relatively high level of apoptosis detected for LP cells activated

5944 in vitro (51, 52). However, we found similar rates and kinetics of T cell apoptosis for LP and spleen in DO11.10 mice activated with OVA323–339 peptide in vivo (based on TUNEL staining of tissue sections; data not shown). Taken together these results suggest that factors present in the tissue enhance LP T cell proliferative responses and survival. Signals delivered from local APC through adhesion molecules (e.g., LFA-1) and costimulatory ligands (CD28, CD40) may reduce LP T cell activation requirements and cell death. It is also possible that soluble factors (e.g., IL-15, IL-7, IL-2, and IL-4) in the LP enhance expansion, lower the activation threshold, and block cell death pathways (53, 54). Taken together, the results of in vivo LP T cell activation suggest that elements of the local microenvironment contribute to potentiating mucosal T cell immune responses. These results suggest that LP CD41 T cells are a unique population of preactivated memory-like cells with the potential to deliver Th1- and Th2-like effector functions. We suspect that both prior activation as well as the mucosal environment combine to increase the level of responses while lowering the threshold of TCR signaling needed for complete stimulation. The relatively high level of cytokines produced by LP T cells may be required to drive highly potent mucosal responses against the type of pathogens encountered at this surface. The reduced activation threshold for stimulation of LP T cells could allow responses to be initiated to relatively low levels of enteric pathogens present early in local infections. Thus, these results suggest that LP CD41 T cells possess a unique set of properties needed to rapidly initiate responses to invasion by life-threatening enteric organisms.

Acknowledgements We thank P. S. Linsley for CTLA4Ig and S. D. Miller for helpful comments on the manuscript.

References 1. Abreu-Martin, M. T., and S. R. Targan. 1996. Regulation of immune responses of the intestinal mucosa. Crit. Rev. Immunol. 16:277. 2. Bloom, P. D., and E. C. Boedeker. 1996. Mucosal immune responses to intestinal bacterial pathogens. Semin. Gastrointestinal Dis. 7:151. 3. James, S. P., C. Fiocchi, A. S. Graeff, and W. Strober. 1986. Phenotypic analysis of lamina propria lymphocytes: predominance of helper-inducer and cytolytic T-cell phenotypes and deficiency of suppressor-inducer phenotypes in Crohn’s disease and control patients. Gastroenterology 91:1483. 4. James, S. P., A. S. Graeff, and M. Zeitz. 1987. Predominance of helper-inducer T cells in mesenteric lymph nodes and intestinal lamina propria of normal nonhuman primates. Cell. Immunol. 107:372. 5. Targan, S. R., R. L. Deem, M. Liu, S. Wang, and A. Nel. 1995. Definition of a lamina propria T cell responsive state: enhanced cytokine responsiveness of T cells stimulated through the CD2 pathway. J. Immunol. 154:664. 6. Senju, M., K. C. Wu, Y. R. Mahida, and D. P. Jewell. 1991. Two-color immunofluorescence and flow cytometric analysis of lamina propria lymphocyte subsets in ulcerative colitis and Crohn’s disease. Digest. Dis. Sci. 36:1453. 7. Schieferdecker, H. L., R. Ullrich, H. Hirseland, and M. Zeitz. 1992. T cell differentiation antigens on lymphocytes in the human intestinal lamina propria. J. Immunol. 149:2816. 8. Schieferdecker, H. L., R. Ullrich, and A. N. Weiss-Breckwoldt. 1990. The HML-1 antigen of intestinal lymphocytes is an activation antigen. J. Immunol. 144:2541. 9. Zeitz, M., H. L. Schieferdecker, S. P. James, and E. O. Riecken. 1990. Special functional features of T-lymphocyte subpopulations in the effector compartment of the intestinal mucosa and their relation to mucosal transformation. Digestion 2:280. 10. Zeitz, M., T. C. Quinn, A. S. Graeff, and S. P. James. 1988. Mucosal T cells provide helper function but do not proliferate when stimulated by specific antigen in lymphogranuloma venereum proctitis in nonhuman primates. Gastroenterology 94:353.

FUNCTIONAL DIFFERENTIATION OF LP T CELLS 11. James, S. P. 1991. Mucosal T-cell function. Gastroenterol. Clin. North Am. 20: 597. 12. Harriman, G. R., E. Hornqvist, and N. Y. Lycke. 1992. Antigen-specific and polyclonal CD41 lamina propria T-cell lines: phenotypic and functional characterization. Immunology 75:66. 13. Swain, S. L. 1995. CD4 T cell development and cytokine polarization: an overview. J. Leukocyte Biol. 57:795. 14. Swain, S. L. 1993. Effector functions of helper T cells. J. Immunother. 14:150. 15. Mosmann, T. R. 1991. Cytokine secretion patterns and cross-regulation of T cell subsets. Immunol. Res. 10:183. 16. Mosmann, T. R., J. H. Schumacher, N. F. Street, R. Budd, A. O’Garra, T. A. Fong, K. W. Moore, A. Sher, and D. F. Fiorentino. 1991. Diversity of cytokine synthesis and function of mouse CD41 T cells. Immunol. Rev. 123:209. 17. Bradley, L. M., K. Yoshimoto, and S. L. Swain. 1995. The cytokines IL-4, IFN-g, and IL-12 regulate the development of subsets of memory effector helper T cells in vitro. J. Immunol. 155:1713. 18. Bradley, L. M., D. D. Duncan, K. Yoshimoto, and S. L. Swain. 1993. Memory effectors: a potent, IL-4-secreting helper T cell population that develops in vivo after restimulation with antigen. J. Immunol. 150:3119. 19. Abbas, A. K., H. J. Burstein, and S. A. Bogen. 1993. Determinants of helper T cell-dependent antibody production. Semin. Immunol. 5:441. 20. Burger, C., and E. S. Vitetta. 1991. The response of B cells in spleen, Peyer’s patches, and lymph nodes to LPS and IL-4. Cell. Immunol. 138:35. 21. Bogen, S. A., I. Fogelman, and A. K. Abbas. 1993. Analysis of IL-2, IL-4, and IFN-g-producing cells in situ during immune responses to protein antigens. J. Immunol. 150:4197. 22. Croft, M., D. D. Duncan, and S. L. Swain. 1992. Response of naive antigenspecific CD41 T cells in vitro: characteristics and antigen-presenting cell requirements. J. Exp. Med. 176:1431. 23. Croft, M., L. M. Bradley, and S. L. Swain. 1994. Naive versus memory CD4 T cell response to antigen: memory cells are less dependent on accessory cell costimulation and can respond to many antigen-presenting cell types including resting B cells. J. Immunol. 152:2675. 24. Panja, A., A. Barone, and L. Mayer. 1994. Stimulation of lamina propria lymphocytes by intestinal epithelial cells: evidence for recognition of nonclassical restriction elements. J. Exp. Med. 179:943. 25. Zeitz, M., H. L. Schieferdecker, R. Ullrich, H. U. Jahn, S. P. James, and E. O. Riecken. 1991. Phenotype and function of lamina propria T lymphocytes. Immunol. Res. 10:199. 26. Murphy, K. M., A. B. Heimberger, and D. Y. Loh. 1990. Induction by antigen of intrathymic apoptosis of CD41 CD81 TCRlo thymocytes in vivo. Science 250:1720. 27. Hurst, S. D., S. M. Sitterding, S. Ji, and T. A. Barrett. 1997. Functional differentiation of T cells in the intestine of T cell receptor transgenic mice. Proc. Natl. Acad. Sci. USA 94:3920. 28. Targan, S. R. 1992. The lamina propria: a dynamic, complex mucosal compartment: an overview. Ann. NY Acad. Sci. 664:61. 29. Saparov, A., C. O. Elson, D. Devore-Carter, R. P. Bucy, and C. T. Weaver. 1997. Single cell analysis of CD41 T cells from ab T cell receptor-transgenic mice: a distinct mucosal cytokine phenotype in the absence of transgene-specific antigen. Eur. J. Immunol. 27:1774. 30. Haskins, K., R. Kobo, J. White, M. Pigeon, J. Kappler, and P. Marrack. 1983. The major histocompatibility complex-restricted antigen receptor on T cells. I. Isolation with a monoclonal antibody. J. Exp. Med. 157:1149. 31. Barrett, T. A., T. F. Gajewski, D. Danielpour, E. B. Chang, K. W. Beagley, and J. A. Bluestone. 1992. Differential function of intestinal intraepithelial lymphocyte subsets. J. Immunol. 149:1124. 32. Linsley, P. S., W. Brady, M. Urnes, L. S. Grosmaire, N. K. Damle, and J. A. Ledbetter. 1991. CTLA-4 is a second receptor for the B cell activation antigen B7. J. Exp. Med. 174:561. 33. Bradley, L. M., S. R. Watson, and S. L. Swain. 1994. Entry of naive CD4 T cells into peripheral lymph nodes requires L-selectin. J. Exp. Med. 180:2401. 34. Carayon, P., and A. Bord. 1992. Identification of DNA-replicating lymphocyte subsets using a new method to label the bromo-deoxyuridine incorporated into the DNA. J. Immunol. Methods 147:225. 35. Tough, D. F., and J. Sprent. 1994. Turnover of naive- and memory-phenotype T cells. J. Exp. Med. 179:1127. 36. Bucy, R. P., A. Panoskaltsis-Mortari, G. Q. Huang, J. Li, L. Karr, M. Ross, K. M. Murphy, and C. T. Weaver. 1994. Heterogeneity of single cell cytokine gene expression in clonal T cell populations. J. Exp. Med. 180:1251. 37. Bucy, R. P., L. Karr, G. Q. Huang, J. Li, D. Carter, K. Honjo, J. A. Lemons, and C. T. Weaver. 1995. Single cell analysis of cytokine gene coexpression during CD41 T-cell phenotype development. Proc. Natl. Acad. Sci. USA 92:7565. 38. Szabo, S. J., N. G. Jacobson, A. S. Dighe, U. Gubler, and K. M. Murphy. 1995. Developmental commitment to the Th2 lineage by extinction of IL-12 signaling. Immunity 2:665. 39. Boirivant, M., I. Fuss, C. Fiocchi, J. S. Klein, S. A. Strong, and W. Strober. 1996. Hypoproliferative human lamina propria T cells retain the capacity to secrete lymphokines when stimulated via CD2/CD28 pathways. Proc. Assoc. Am. Physicians 108:55. 40. Kanamori, Y., K. Ishimaru, M. Nanno, K. Maki, K. Ikuta, H. Nariuchi, and H. Ishikawa. 1996. Identification of novel lymphoid tissues in murine intestinal mucosa where clusters of c-kit1 IL-7R1 Thy11 lympho-hemopoietic progenitors develop. J. Exp. Med. 184:1449. 41. Sanderson, I. R., A. J. Ouellette, E. A. Carter, W. A. Walker, and P. R. Harmatz. 1993. Differential regulation of B7 mRNA in enterocytes and lymphoid cells. Immunology 79:434.

The Journal of Immunology 42. Bland, P. W., and D. M. Kambarage. 1991. Antigen handling by the epithelium and lamina propria macrophages. Gastroenterol. Clin. North Am. 20:577. 43. Panja, A., R. S. Blumberg, S. P. Balk, and L. Mayer. 1993. CD1d is involved in T cell-intestinal epithelial cell interactions. J. Exp. Med. 178:1115. 44. Abe, R., P. Vandenberghe, N. Craighead, D. S. Smoot, K. P. Lee, and C. H. June. 1995. Distinct signal transduction in mouse CD41 and CD81 splenic T cells after CD28 receptor ligation. J. Immunol. 154:985. 45. Cunningham, A. C., and J. A. Kirby. 1995. Regulation and function of adhesion molecule expression by human alveolar epithelial cells. Immunology 86:279. 46. Tian, L., R. J. Noelle, and D. A. Lawrence. 1995. Activated T cells enhance nitric oxide production by murine splenic macrophages through gp39 and LFA-1. Eur. J. Immunol. 25:306. 47. Yamamoto, N., J. P. Zou, X. F. Li, H. Takenaka, S. Noda, T. Fujii, S. Ono, Y. Kabayashi, N. Mukaida, K. Matsushima, et al. 1995. Regulatory mechanisms for production of IFN-g and TNF by antitumor T cells or macrophages in the tumor-bearing state. J. Immunol. 154:2281. 48. Kuhn, R., J. Lohler, D. Rennick, K. Rajewsky, and W. Muller. 1993. Interleukin10-deficient mice develop chronic enterocolitis. Cell 75:263.

5945 49. Davidson, N. J., D. J. Berg, M. Leach, and D. M. Rennick. 1995. CD41 lamina propria lymphocytes from IL-102/2 mice can mediate enterocolitis in RAG-22/2 recipient mice. Clin. Immunol. Immunopathol. 76:544 (Abstr. 259-0). 50. Croft, M., and S. L. Swain. 1995. Recently activated naive CD4 T cells can help resting B cells, and can produce sufficient autocrine IL-4 to drive differentiation to secretion of T helper 2-type cytokines. J. Immunol. 154:4269. 51. Boirivant, M., R. Pica, R. DeMaria, R. Testi, F. Pallone, and W. Strober. 1996. Stimulated human lamina propria T cells manifest enhanced Fas-mediated apoptosis. J. Clin. Invest. 98:2616. 52. De Maria, R., M. Boirivant, M. G. Cifone, P. Roncaioli, M. Hahne, J. Tschopp, F. Pallone, A. Santoni, and R. Testi. 1996. Functional expression of Fas and Fas ligand on human gut lamina propria T lymphocytes: a potential role for the acidic sphingomyelinase pathway in normal immunoregulation. J. Clin. Invest. 97:316. 53. Lodolce, J. P., D. L. Boone, S. Chai, R. E. Swain, T. Dassopoulos, S. Trettin, and A. Ma. 1998. IL-15 receptor maintains lymphoid homeostasis my supporting lymphocyte homing and proliferation. Immunity 9:669. 54. Boise, L. H., and C. B. Thompson. 1996. Hierarchical control of lymphocyte survival. Science 274:67.