Human Airway Epithelial Cells Stimulate T-Lymphocyte ... - ATS Journals

Human Airway Epithelial Cells Stimulate T-Lymphocyte Lck and Fyn Tyrosine Kinase Thomas H. Kalb, Xian Yang Yio, and Lloyd Mayer Divisions of Clinical Immunology and Pulmonary/Critical Care, Department of Medicine, Mt. Sinai Medical Center, New York, New York

Previous studies have shown that human airway epithelial cells (AEC) can stimulate allogeneic peripheral blood T-lymphocyte (PBT) proliferation. We now sought to determine which AEC surface molecule/ T-cell coreceptors are involved in this process. AEC-induced PBT proliferation was inhibited by 25 mM genestein or herbamycin A (0.9 and 1.8 mM), both tyrosine kinase inhibitors. Anti-phosphotyrosine immunoblots performed on PBT lysates after coculture with AEC demonstrated phosphorylation of 56kD and 60kD bands. To determine whether CD3 associated p59fyn, or CD4 and CD8 associated p56lck phosphotyrosine kinases (PTK) were involved, we assayed kinase activity in lymphocyte lysates immunoprecipitated with anti-p56lck and p59fyn mAbs. PBT cells or murine T-cell line transfectants expressing human CD4 (3G4) or human CD8a (3G8) were cocultured with AEC or A549, an alveolar-like cell line lacking class II Ag expression. After A549 or AEC coculture, p56lck activity in PB T-cells peaked at 2 min whereas p59fyn kinase activity continued to rise at 8 min. AEC (expressing class II Ags) stimulate PTK activity in both 3G8 and 3G4 cells. A549 stimulated p56lck in 3G8, but not in 3G4 cells. This activation of p56lck was not blocked by preincubation of A549 with anti-class I or anti-CD1d mAbs. An antibody generated in our laboratory, which recognizes an epithelial specific surface molecule (mAb L12) and which blocks AEC driven PBT proliferation, was shown to block PTK activity of peripheral blood T-cell lysates, though not of 3G8 lysates. These studies suggest that AEC are capable of stimulating CD4 and CD8 associated lck and CD3 associated fyn kinases through class II dependent and independent pathways. Kalb, T. H., X. Y. Yio, and L. Mayer. 1997. Human airway epithelial cells stimulate T-lymphocyte lck and fyn tyrosine kinase. Am. J. Respir. Cell Mol. Biol. 17:561–570.

Mature T-lymphocytes become activated to perform their effector functions when stimulated by appropriate antigen presenting cells (APCs) bearing MHC class I or class II molecules. These surface molecules bind domains on the T-cell receptor (TcR) and simultaneously engage either CD8 (for class I) or CD4 (for class II) (1, 2). Human airway epithelial cells (AEC) and intestinal epithelial cells (IEC) resemble conventional antigen presenting cells (APC) such as dendritic cells in that they express class II MHC molecules, and stimulate peripheral blood T-lymphocyte proliferation in MLR cultures (3–6). Airway epithelial

(Received in original form January 11, 1996 and in revised form January 22, 1997) Address correspondence to: Thomas H. Kalb, M.D., Division of Pulmonary/Critical Care, Box 1232, Mount Sinai Medical Center, 1 Gustave Levy Place, New York, NY 10029. Abbreviations: Human bronchial airway epithelial cells, AEC; antigen, Ag; Dulbecco’s Modified Essential Medium, DMEM; major histocompatibility complex, MHC; mixed lymphocyte reaction, MLR; phytohemaglutinin, PHA; phosphotyrosine kinase, PTK; T lymphocyte antigen receptor, TcR. Am. J. Respir. Cell Mol. Biol. Vol. 17, pp. 561–570, 1997

cells have also been shown to support mitogen and antigen specific T-lymphocyte proliferation, though the response appears to be relatively weak compared with conventional accessory cells (7, 8). In epithelial cell-stimulated MLR cultures, T-cell proliferation can be blocked by preincubating AEC, but not IEC, with an anti-class II mAb. These findings suggest that AEC, like conventional APC, express functional class II molecules which can be recognized by alloreactive T-cells. An airway derived epithelial cell line which is devoid of class II Ag expression (A549) is incapable of stimulating T-cell proliferation (9). Although these observations suggest that class II molecules on AEC directly engage TCR on allogeneic T-lymphocytes, soluble mediators of epithelial origin may stimulate or enhance T-cell proliferation or the expansion of a subset of preactivated T-cells. Airway epithelial cells have been shown to elaborate a battery of mediators which may affect T-cell function including IL-6, IL-10, IL-11, IL-16, GM-CSF, MIP1a, and PGE2 (9–17). Alternatively, class II Ag bearing cells other than AEC may contribute to lymphocyte proliferation, since small numbers of class II Ag expressing dendritic cells or airway derived T cells may contaminate these MLR cultures, de-

562

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 17 1997

spite preculture techniques aimed at excluding these cell types from AEC isolates (18, 19). Distinguishing among these possible causes of T-cell proliferation in AEC stimulated MLR may be important, since if AEC function as APCs in vivo, then they may provide a unique set of signals to airway T-cells with distinct functional consequences. The earliest detectable biochemical event after stimulation of T-lymphocytes through their antigen receptor (TcR) is activation of a set of cytoplasmic tyrosine kinases (PTK), and subsequent phosphorylation of a number of substrates (20, 21). Two related src family tyrosine kinases have been implicated in these early events: CD3 associated p59fyn kinase, and p56lck kinase, which can be associated with either CD4 or CD8a cytoplasmic domains (23–26). In this study, we sought to clarify these potential activation pathways induced by AEC/T-cell contact as a way to determine the T-cell surface molecules which could be engaged by counterreceptors on AEC. The capacity of airway epithelial cells to stimulate T-lymphocytes was assessed by examining PTK activity induced by airway cell/ T-cell coculture. We found that AEC induced T-cell proliferation depends on PTK activity, similar to the effects mediated by IEC and conventional APCs. AEC dependent PTK activation was rapid, and could be demonstrated in the absence of contaminating class II bearing non-epithelial cells. Airway epithelial cells and cell lines stimulated peripheral blood T-cells via both p59fyn kinase associated with CD3/ TcR, and p56lck associated with either CD4 or CD8 a, though class II antigen expression on AEC is required for PTK activation of a CD4 1 T-cell line. In contrast, CD81 T-cell stimulation may be induced by airway cells independent of class I or class II molecules. This latter finding may be explained by a novel epithelial cell specific antigen which can be recognized by T cells and which is involved in PTK activation in T cells.

Materials and Methods Cell Isolation and Cell Lines Murine T-cell transfectants 3G4 and 3G8, transfected with a full length human CD4 or a full length CD8a cDNA, respectively, were kind gifts from Dr. S. J. Burakoff (DanaFarber Institute, Boston, MA). The construction of these transfectants have been previously described (26, 27). The only human proteins expressed on these cells are either CD4 or CD8 (aa homodimers) and others have previously documented that murine p56lck can associate with human CD4 or CD8. Crosslinking CD4 with anti-CD4 mAbs on 3G4 cells or crosslinking CD8 with anti-CD8 mAbs on 3G8 cells results in the autophosphorylation of p56lck and phosphorylation of a number of other substrates. These cells were maintained at 378C, 5% CO2 in RPMI 1640 supplemented with 10% fetal calf serum (GIBCO, Grand Island, NY), glutamine 2 mM (GIBCO), penicillin 50 mg/ml/streptomycin 50 mg/ml (GIBCO), referred to as culture medium (CM). The A549 cell line was used to address the requirement for class II molecules in epithelial cell stimulated T-cell activation. These adherent cells were maintained in Ham’s F12 supplemented with 10% FCS (GIBCO), gluta-

mine (2 mM), penicillin/streptomycin (50 mg/ml) (GIBCO) at 378C in a 5% CO2 humidified incubator (Hotpack, Philadelphia, PA). LA-1, a thyroid derived epithelial cell line, and DLD-1, an intestinal epithelial cell line, were used to compare the stimulation induced by A549; these lines were maintained in CM. Human airway epithelial cells (AEC) were isolated from thoracotomy specimens and transported in cold (48C) Dulbecco’s Modified Essential Medium (DMEM; BioWhittaker, Walkersville, MD) with penicillin/streptomycin (50 mg/ml) as previously described (3). Briefly, lobar and segmental bronchi which are grossly free of disease, were dissected and trimmed free of adjoining tissue, and incubated at 378C with Dispase (3 mg /ml; Boehringer Mannheim, Germany) in DMEM for 30 min. Airway mucosae were exposed and scraped with a scalpel, and cell suspensions were filtered through nylon mesh filter (53 mm mesh; Spectrum, Los Angeles, CA). Epithelial cell isolates were resuspended in serum-free growth factor supplemented medium adapted from the preparation methods described by Reen Wu and colleagues (28), consisting of Ham’s F12 (GIBCO) supplemented with 50 mg/ml penicillin/streptomycin (GIBCO), hydrocortisone (1 mM) (Collaborative Research Inc., Bedford, MA), transferrin (5 mg/ml) (Sigma Chemical, St. Louis, MO), insulin (5 mg /ml) (Collaborative Research), epidermal growth factor (25 ng/ml) (Sigma), and bovine hypothalamic extract (10 mg /ml) (Collaborative Research), referred to as F12 1 5F. Cells were plated onto round-bottom microwell plates (Flow Laboratories, MacLean, VA) at a density of 105/0.2 ml. Medium was exchanged every two days. T-lymphocytes utilized in MLR were obtained by rosetting peripheral blood lymphocytes (obtained by ficoll hypaque density centrifugation) with neuraminidase treated sheep red blood cells. T-cell preparations were . 95% CD31 and , 2% CD141 (by flow cytometry). To further minimize contamination by accessory cells, rosetted fractions were subjected to nylon wool column separation as previously described (30), reducing the CD201 and CD141 accessory cell population to , 1%. Depletion of monocytes was confirmed by a loss of mitogen responsiveness. Immunoblot with anti-phosphotyrosine mAb. 0.5 3 106 peripheral blood T cells or 3G4 or 3G8 cells were cocultured with 2.5 3 105 A549 or human bronchial epithelial cells, for varying periods (0, 0.25, 0.5, 1, 2, 5, 8 min) in 0.5 ml eppendorf tubes maintained at 378C. Crosslinking with anti-CD4 (OKT4) or anti-CD8 (OKT8) monoclonal antibodies (mAbs) and rabbit anti-mouse Ig (Organon Teknika, Durham, NC) served as positive controls. Where indicated, airway cells were preincubated with mAb for 30 min at 378C prior to coculture with T cells: W6/32 (anticlass I mAb obtained from ATCC, Rockville, MD); VG2.2 (anti-class II MHC, monomorphic determinant, a kind gift of Dr. Shu Man Fu, University of Virginia, Charlottesville, VA); 3C11, anti-CD1d (a kind gift of Rick Blumberg, Brigham and Women’s Hospital, Boston, MA); mAb L12, an anti-human IEC membrane antigen antibody generated in our laboratory which has been previously described (30). All cells were washed vigorously in serum free RPMI at 378C prior to coculture or addition of crosslinking antibody. After a brief incubation, the T cells were

Kalb, Yio, and Mayer: Human AEC Stimulate T-lymphocyte Tyrosine Kinases

lysed in ice-cold hypotonic buffer (0.18% phosphate buffered saline [PBS]) containing phosphatase and protease inhibitors (200 mM Na3VO4, 1 mM PMSF, 20 mg/ml Aprotinin, 5 mM iodoacetamide, 20 mg/ml leupeptin) (Sigma Chemical Co.) and lysates were resolved on 10% SDSPAGE and transferred onto a nitrocellulose membrane (Schleicher & Schuell, Inc., Keene, NH) in transfer buffer (20% methanol, 150 mM glycine, 25 mM Tris, pH 8.3). After transfer, the nitrocellulose membrane was blocked by 100 ml of 5% nonfat milk in PBS, and incubated for 2 h at room temperature with the anti-phosphotyrosine mAb 4G10 (1 mg/ml; Upstate Biotechnology Inc., Lake Placid, NY) in 0.5% non-fat milk in phosphate buffered saline (PBS). After washing 5 times in washing buffer (0.05% Tween 20 in PBS), secondary antibody was added for 2 h (2 mg/ml, horseradish peroxidase conjugated goat antimouse IgG; Cappel Laboratory, Durham, NC). After washing 5 times with washing buffer, phosphorylated tyrosine residues were identified by a chemiluminescence reagent (Dupont ECL, Wilmington, DE). Immunoprecipitation and kinase assay. Precleared lysates from cell cocultures (incubated with protein A sepharose beads for 1 h) were precipitated with either anti-lck (UBI), or anti-fyn Ab (heterologous rabbit antisera) (Santa Cruz Biotech, Santa Cruz, CA) conjugated protein A sepharose beads. Washed beads (PBS followed by LiCl 10 mM, Tris 20 mM, pH 8) were incubated in kinase buffer (MnCl2 10 mM, Tris 50 mM, pH 7.4) with gP32-ATP for 30 min at room temperature. Reduced and boiled samples were analyzed by SDS-PAGE and autoradiography to resolve autophosphorylated p56lck or p59fyn and associated substrates. Mixed lymphocyte/epithelial cell cultures. Allogeneic mixed cell cultures were performed using 5 3 104 irradiated (3,000 rads from a cesium source) epithelial cells as stimulators and 1 3 105 allogeneic T-cells as responder cells, in RPMI with 5% agammaglobulinemic serum. Maximal stimulation of the T-cell population was determined by coculture of T cells with PHA (1 mg/ml) (GIBCO). Where indicated, genestein (a PTK inhibitor) (Calbiochem-Novabiochemical Corp., La Jolla, CA) (25 mM) was added to isolated T-lymphocytes for 2 h at room temperature, and washed vigorously in CM immediately before coincubation with epithelial cells. In other experiments, inhibition by genestein was compared with herbamycin, a selective tyrosine kinase inhibitor, added in varying concentrations (0.45–1.8 mM) at the onset of culture. All cultures were performed in triplicate in 96 well round bottom microwell plates (Flow Laboratories) for 120 h at 378C in a 5% CO2 humidified incubator. Stimulation was determined by the incorporation of [3H]thymidine (1 mCi/well; 2 Ci /mmol) (ICN, Irvine, CA) added for the final 18 h of culture. Mean, standard deviation, and statistical significance (using Student’s t test) were performed with a Statworks software package (Apple Corporation, Cupertino, CA).

Results Inhibition of AEC Stimulated T-cell Proliferation in the Presence of Tyrosine Kinase Inhibitors Previous studies from our lab have demonstrated that both intestinal and airway epithelial cells can stimulate al-

563

logeneic T-cell proliferation. IEC appear to selectively stimulate CD81 T-cells via activation of CD8a associated p56lck. In contrast, AEC, like conventional antigen presenting cells, preferentially activate CD41 T-cell proliferation. To determine whether there is a similar dependence upon kinase activation in this system, we cocultured AEC with allogeneic T-cells in the presence or absence of genestein, a tyrosine kinase inhibitor (31). As seen in Figure 1a, the ability of AEC to induce allogeneic T-lymphocyte proliferation in MLR culture, as detected by [3H]thymidine uptake, was ablated by the preincubation of lymphocytes with genestein at the 25 micromolar concentration used in these studies. Addition of genestein 30 min after T cells were cocultured with airway cells was no longer effective in blocking proliferation, suggesting that PTK activation is

Figure 1. Inhibition of AEC-stimulated T-cell proliferation in the presence of the protein tyrosine kinase inhibitors. (a) T cells treated or untreated with genestein (25 mM) 2 h before or 30 min after the onset of culture were incubated with irradiated AEC for 5 days. [3H]thymidine was added 18 h before cell harvesting, and thymidine incorporation was measured in a scintillation counter. Genestein (at concentrations , 150 mM) had no inhibitory effect on T cells stimulated with PHA (T 1 PHA 5 103,040 cpm, T 1 PHA 1 150 mM genestein 5 108,350 cpm, T 1 PHA 1 25 mM genestein 5 112,266 cpm). These findings are representative of four experiments. (b) T cells treated or untreated with Herbamycin A (0.45–1.8 mM) at the onset of culture with irradiated AEC for 5 days, or with the addition of PHA 1%.

564


a necessary early event, but is not required for downstream maintenance of proliferation. Furthermore, this finding suggests that the effect of genestein is specific, and not related to toxicity. Also attesting to the lack of toxicity, T-lymphocyte proliferative response to phytohemaglutinin [PHA 1 mg/ml] was unaffected by genestein at concentrations less than 150 mM (see Figure 1 legend). We confirmed these data assessing the effects of herbamycin, a more selective tyrosine kinase inhibitor. As seen in Figure 1b, herbamycin (at varying concentrations) added at the onset of culture inhibited AEC-driven T-cell proliferation. These findings support the concept that activation of a tyrosine kinase is critical to the T-cell proliferation induced by AEC. Anti-phosphotyrosine Immunoblots Since tyrosine kinase activity is required in this system, we next assessed which kinases were activated in these cocultures. Normal AEC were cocultured with allogeneic peripheral blood T-lymphocytes for one minute and antiphosphotyrosine immunoblots were performed (Figure 2a). Multiple tyrosine phosphorylated bands were induced, including a 56 kD band, presumably p56lck, as well as a 60 kD band molecule, presumably fyn kinase. We also asked which specific surface molecules on AEC were involved in this tyrosine kinase activation. In previous proliferation studies, we have found differences in the inhibitory effect of antibodies directed at epithelial surface molecules in the stimulation induced by airway versus intestinal epithelial cells: AEC induced proliferation of T cells is inhibited by anti-class II mAb, similar to conventional accessory cells such as monocytes, whereas intestinal epithelial cell stimulation is not. Conversely, intestinal epithelial cell stimulation is ablated by the anti-nonclassical class I molecule antibody, 3C11, whereas airway cell stimulation of T cells is unaffected by such treatment (32). We have hypothesized that these antibodies inhibit cell/cell interactions involving receptor/ligand pairing, and that this interaction is required for proliferation. In the present study, we asked whether AEC-induced tyrosine kinase activity was affected by antiepithelial surface molecule antibodies which inhibit proliferation. In the gut system, p56lck tyrosine kinase activation by intestinal epithelial cells appears to be independent of class II and class I MHC (33). In the airway system, AEC-induced PTK activation is inhibited by an anti-class II mAb VG2.2 (Figure 2a, lane labeled MHC II), but is unaffected by an anti-class I mAb w6/32 (MHC I), or an anti-class Ib antibody 3C11 (CD1d), as compared with the isotype control antibody (IgG). These effects parallel the effects of these Abs on the proliferation response. In the intestinal system, recent data suggest that CD1d may coassociate with a novel surface molecule expressed on intestinal epithelial cells, gp180, and that gp180 may serve as a novel CD8 ligand (30). gp180 is recognized by two mAb, B9 and L12, which recognize nonoverlapping epitopes. AEC and A549 cells do not express the molecule recognized by mAb B9 but can be stained with mAb L12. Therefore, it is plausible that an altered form of gp180 exists on airway cells which is still involved in the activation

Figure 2. Anti-phosphotyrosine immunoblot of lysates from allogeneic peripheral blood T-lymphocytes cocultured with AEC. (a) Allogeneic peripheral blood T-lymphocytes were cocultured with AEC (AEC 1 PB T 19), with anti-CD-4 mAb OKT4 followed by crosslinking heterologous rabbit anti-mouse IgG Ab (CD4), or with anti-CD8 mAb followed by RAM (CD8), lysed, and resolved on 10% SDS-PAGE. AEC were preincubated with indicated mAb for 1 h at 378C prior to coculture (IgG, MHCI, CD1d, MHCII, L12). One minute time point refers to elapsed coculture or RAM incubation for 1 min at 378C prior to the addition of icecold lysis buffer. Control lane To 1 Ao refers to no coincubation prior to the addition of cold lysis buffer to PB T and AEC. The gel was transferred onto nitrocellulose and subjected to Western blot analysis using the antiphosphotyrosine mAb 4G10 as described in MATERIALS AND METHODS. This blot is representative of at least three experiments. Lysed AEC (without T cells) did not constitutively express any tyrosine phosphorylated bands. (b) Effect of L12 mAb preincubation on AEC stimulated allogeneic T-cell proliferation. 10 5 allogeneic peripheral blood T-lymphocytes were incubated with irradiated AEC (5 3 10 4) for 5 days. Where indicated, AEC were preincubated with mAb L12 or anti-class II mAb (VG2.2) for 2 h followed by washing and coculture with T cells. [3H]thymidine was added 18 h before cell harvesting, and thymidine incorporation was measured in a scintillation counter. These findings are representative of four experiments.

of CD8-associated p56lck. In support of this hypothesis, L12 mAb preincubation of AEC significantly inhibits an AEC stimulated MLR (Figure 2b; T cells 1 AEC 1 control IgG1 12,908 6 5,327 cpm; T cells 1 AEC 1 L12 mAb 1,395 6 682 [P , 0.05]), whereas an anti-class I mAb (W6/ 32) has minimal or no inhibitory effect. As seen in Figure 2a, preincubation of AEC with mAb L12 prior to cocul-


ture with T cells resulted in diminished tyrosine phosphorylation (lane L12). Thus, there appears to be a non-MHC surface antigen expressed on AEC which is involved in PTK activation. Once again, inhibition of the PTK activation correlates with inhibition of proliferation. Since AEC preparations could potentially be contaminated with other accessory cell populations, we attempted to confirm our results using the cell line A549, a class II Ag negative cell line derived from airway epithelium. This also allowed us to further assess the requirement for class II Ag on airway epithelial cells acting as accessory cell stimulators, since the antibody inhibition study suggests that class II MHC plays a pivotal role in AEC induced T-cell stimulation. Interestingly, we found that A549 was capable of inducing tyrosine phosphorylation of a 56 kD band, presumably native p56lck, as well as a 60 kD band, peaking between 1 and 2 min in periperal blood T cells (Figure 3), similar to the phosphorylation pattern seen when AEC from human airway were used. Thus airway epithelial cells appear to stimulate T-cell phosphotyrosine kinases through class II independent as well as dependent pathways. Although gentle hypotonic lysis conditions were used to limit epithelial cell lysis, A549 cells did contribute variably one or more phosphorylated bands to the PTK immunoblots, and control lanes of isolated A549 cells were included in all immunoblots to control for these bands.

565

family of similar size (e.g., p53/p56 lyn, p56blk, or p62cyes) have been described to play roles in B lymphocyte, platelet and mast cell activation respectively, but have not been noted to transduce and amplify exogenous signals received by T cells or T-lymphocyte lines (34, 35). Lck and Fyn activity was assessed in immunoprecipitates prepared in parallel from the same peripheral blood T-lymphocyte lysates after incubation with A549 cells. These studies demonstrated kinase activation of both p56lck as well as p59fyn. Whereas lck immunoprecipitates showed peak kinase activity at 2 min (Figure 4a), fyn kinase activity was detectable at 1 min, though it continued to rise at 8

Kinase Assay—Detection of Phosphotyrosine Kinase Activity After Lck or Fyn Immunoprecipitation To determine whether the phosphotyrosine kinase activity detected in these assays was due to either p56lck or p59fyn kinase activation, we next performed kinase assays on anti-lck (Figure 4a) and anti-fyn (Figure 4b) immunoprecipitates in lysates of cocultured cells. These were considered the most likely candidates for the observed PTK activity, based on their well characterized role in conventional T-lymphocyte activation and the molecular weights of the observed bands. Other members of the cytoplasmic PTK

Figure 3. Anti-phosphotyrosine immunoblot: A549 cocultured with peripheral blood T-lymphocytes. 0.5 3 10 6 allogeneic peripheral blood T-lymphocytes were cocultured alone (T), with antiCD8 mAb 1 RAM for 1 min (anti-CD8), or with 2.5 3 10 6 A549 cells for varying time points, lysed and resolved on 10% SDSPAGE. Western blot analysis was performed using the antiphosphotyrosine mAb 4G10 as described in M ATERIALS AND METHODS. This blot is representative of at least three experiments.

Figure 4. Comparison of kinetics of lck and fyn kinase activation induced by A549 cells. Lysates of peripheral blood T-lymphocytes alone or after incubation with either anti-CD4 mAb or A549 cells were precipitated with anti-lck mAb, or anti-fyn mAb conjugated protein A sepharose beads. Washed beads were incubated with gP32-ATP for 30 min at room temperature. Samples were analyzed by SDS-PAGE and autoradiography to resolve autophosphorylated p56lck or p59fyn and associated precipitates. (a) Antilck immunoprecipitates; (b) anti-fyn immunoprecipitates. These blots are representative of at least three experiments.

566


min as suggested by increased intensity of the 59 kD band (Figure 4b). In addition to the phosphorylated band seen at 56 kD and 59 kD, other phosphorylated bands were detected in these lysates which were either absent or of much lower intensity in control lanes. These bands are consistent with coprecipitates known to associate with activated lck or fyn, and may be phosphorylated by activated kinase in both peripheral lymphocytes and their murine counterparts. Several such coprecipitates have been previously identified and may serve as downstream mediators of PTK activity, such as the GAP-associated p62, the p62-related 68 kD protein, and hnRNP K (36). To dissect out the potential AEC/ T-cell binding partners which are involved in PTK activation, we utilized a model system where a murine T-cell hybridoma had been transfected with either full length human CD4 cDNA (3G4) or full length human CD8a cDNA (3G8). In this system, the only human proteins expressed are either CD4 or CD8 which can associate with murine p56lck. Furthermore, each cell line serves as a nonspecific control for the other. Coculture of A549 with 3G8 resulted in the phosphorylation of a 56 kD band and associated substrates, whereas no such activation was detected in 3G4 cells (Figure 5). Thus, an AEC cell line with no detectable surface class II molecules was capable of stimulating CD8 associated PTK activity with an efficiency equal to that of freshly isolated human AEC, but unlike AEC, could not activate CD4-associated PTK (see below). Since class I MHC molecules are the conventional ligand for CD8, we considered the possibility that epithelial cell class I Ag expression was sufficient for activation of CD8-associated p56lck. However, a thyroid derived epithelial cell line, LA-1, which expresses class I Ag at a level comparable to A549 and an intestinal cell line DLD-1, did not induce PTK activation in 3G8 (or 3G4) (Figure 6b). In contrast, PTK activation in 3G8 cells was induced by DLD-1, an intestinal epithelial cell line expressing both class I as well as Ags recognized by mAbs B9 and L12. These results suggest that class I Ag expression is not sufficient to stimulate CD8 associated p56lck, and support the hypothesis that non-class I molecules expressed on A549 and DLD-1 are capable of engaging CD8 and inducing PTK activation, perhaps related to the expression of molecules recognized by the L12 mAb.

Figure 5. A549 stimulates CD8a, but not CD4 associated PTK activity. A549 cells (5 3 10 5) were cocultured with 3G4 or 3G8 cells (10 6 ) and antiphosphotyrosine immunoblots were performed on cell lysates. This blot is representative of at least three experiments.

We next contrasted AEC with A549 induction of PTK activity in 3G4 and 3G8. As seen in Figure 6a, coculture of AEC with either 3G4 or 3G8 resulted in the detection of both a 56 and 60 kD tyrosine phosphorylated band (3G8 . 3G4). Thus, both AEC and A549 express a non-class I molecule capable of activating CD8-associated kinases, but only AEC were shown to stimulate CD4 associated PTK. Since the expression of class II Ag by the epithelial cells correlates with PTK activation in 3G4 cells, we speculate that stimulation of CD41 T-cells depends upon class II Ag expression by AEC. Anti-lck Immunoprecipitates Anti-epithelial cell mAb L12 blocks AEC induced lck activation. To more directly address the possibility that inhibition of tyrosine phosphorylation by L12 mAb observed in immunoblots occured through inhibition of p56lck activation, kinase activation was assayed in anti-lck immunopre-

Figure 6. (a) AEC stimulate both CD4 and CD8a associated PTK activity. 3G4 or 3G8 cells (10 6) were cocultured with freshly harvested human airway epithelial cells (AEC) (2.5 3 10 5), and antiphosphotyrosine immunoblots were performed (mAb 4G10) with cell lysates. This blot is representative of at least three experiments. (b) Comparison of PTK stimulation induced by epithelial cell lines of thyroid and intestinal origin. LA-1a thyroid derived cell line, or DLD-1 intestinal cell line, were incubated with 3G8 or 3G4 cell lines for varying intervals, and PTK induction in cell lysates was determined by antiphosphotyrosine immunoblot.


cipitates of coculture lysates where isolated human AEC were preincubated for 1 h with L12 mAb or control Ab. In previous studies with intestinal epithelial cells, mAb B9, which recognizes gp180, was shown to inhibit lck activation. As described above, AEC and A549 cells do not express the molecule detected with mAb B9 but are stained with mAb L12. B9 Ab had no effect on AECinduced PTK activation, or lck kinase activity. Thus, it remains unclear whether the molecule recognized by L12 mAb on AEC is functionally related to gp180, or whether this molecule is involved in the activation of CD8-associated p56lck. In Figure 7a, it can be seen that preincubation of AEC with mAb L12 resulted in diminution of lck kinase activity in peripheral blood T-cell/AEC coculture, as seen by diminished intensity of autoradiography at the 56 kD band, compared with either CD8 crosslinking or stimulation by AEC preincubated with an isotype control antibody. However, no diminution of p56 kinase activity was noted in 3G8 cells (Figure 7b). Since CD8 on 3G8 cells is the only human ligand available for binding by AEC, this study supports a role for the CD8 molecule in the stimulation observed in peripheral blood T cells. However, the difference in the effect of L12 in diminishing PTK activity in peripheral blood cells but not in 3G8 cells suggests that L12 is not blocking a ligand for CD8. The effect of L12 to diminish proliferation, PTK activation, and lck kinase activity in AEC/T cell cocultures may be mediated by another ligand binding pair which indirectly affects activation of CD8-associated lck. This difference between the action of L12 mAb on AEC and B9 mAb on IEC may relate to different ligand specificities. We next sought to analyze the manner by which fyn kinase is activated by AEC. Fyn kinase is reported to be activated after crosslinking of TcR or CD3 components. Sequential lck and fyn activation has also been reported to follow crosslinking ICAM-3 on Jurkat T-cell line, and it was suggested that fyn may serve as a substrate for activated lck in this system. Therefore, activation of fyn kinase seen in Figure 4 suggests either that the TcR/CD3 complex recognizes an A549 surface molecule, or that fyn serves as a substrate for activated lck kinase, leading to secondary activation of fyn kinase. This alternative would be consistent with the observed time course of activation, with sequential phosphorylation of lck followed by fyn in the above noted experiment. To test for the possibility of indirect activation of fyn by lck, immunoprecipitation and kinase assays for both lck and fyn were performed after A549 was incubated with either 3G8 or 3G4 cells. Since direct stimulation of these cell lines via surface molecule engagement only can occur through CD8 (the murine TcR would not be expected to recognize human restriction elements), phosphorylation of fyn kinase in this experiment would imply an indirect effect, presumably through CD8-associated lck kinase activity, consistent with the antiphosphotyrosine blots. Lck was activated in 3G8 (Figure 8), but not in 3G4 cells. However, unlike the stimulation of fyn in peripheral blood T-cells by A549, no phosphorylation of fyn was seen in either 3G8 or 3G4 when stimulated by A549. Thus, indirect phosphorylation of fyn kinase is not likely to occur in this system.

567

Figure 7. Antiepithelial cell mAb L12 blocks AEC induced lck activation. Lysates from AEC cell coculture with peripheral T cells (a) or 3G8 cell line (b) were precipitated with anti-lck mAb conjugated protein A sepharose beads. AEC were preincubated with mAb L12 or isotype control for 1 h at 37 8C prior to coculture. Control lanes include lymphocytes alone (T), lymphocytes incubated with RAM 19 (RAM), lymphocytes incubated with antiCD8 mAb 1 h, followed by RAM 19 (CD8 1 RAM). Washed beads were incubated with gP32-ATP for 30 min at room temperature, boiled with reduced sample buffer, and analyzed by SDSPAGE and autoradiography to resolve autophosphorylated p56lck and associated precipitates.

Discussion In these studies, the capacity of airway epithelial cells to stimulate T-lymphocytes was assessed by examining phosphotyrosine kinase activity induced by airway cell /T-cell coculture. These studies clarified several aspects of AEC induced T-cell proliferation observed in mixed cell cultures. First, PTK activation is a very early event after AEC/T-cell contact (within seconds to minutes), compared with the induction of proliferation which takes days in mixed cell cultures. By detection of this early event, the PTK assay strongly implicates a contact dependent effect by epithelial cell surface molecules. In this brief time frame, an effect of

568


Figure 8. Fyn kinase is not indirectly phosphorylated after activation of lck kinase by A549. A549 was incubated with either 3G8 or 3G4 cells, expressing transfected CD8a or CD4 molecules respectively, and anti-lck and anti-fyn immunoprecipitation and kinase assays were performed.

soluble mediators is likely to be negligible—both airway and T cells were washed vigorously immediately prior to coincubation. Second, since clonal populations of stimulator and responder cells were used, our assays were free of nonepithelial accessory cells. Thus, the observed PTK activation in these studies must depend on a direct effect of airway cells. Since AEC from thoracotomy specimens induced similar patterns of lymphocyte PTK stimulation, these effects are most likely attributed to AEC rather than another accessory cell type which may be included in the stimulator population. By extension, these experiments support the interpretation that AEC are capable of accessory cell function and are responsible for the T-cell proliferation observed in MLR, independent of other accessory cell populations. Third, AEC surface molecules appear to stimulate signaling pathways in T cells by engaging TCR /CD3 and either CD4 or CD8. This is suggested by the activation of PTKs which are specifically associated with these surface molecules. Although p56lck can also associate with CD2, IL-2R, and ICAM-3 (37–39), only human CD4 or CD8 is present on transfected cell lines (3G4 and 3G8, respectively), limiting the binding partners available for AECs in experiments in which these cell lines were used. T-cell signaling is highly plastic, complex, and variable. Identification of p56lck or p59fyn activation by these experients probes a narrow, isolated aspect of this process. Although useful to suggest surface molecule engagement by airway cells, it is clear that PTK activation does not translate into proliferation, as demonstrated by the lack of T-cell proliferation in A549 stimulated MLR despite PTK activation. Depending on costimulation, stage of cell cycle, soluble mediators and other variables, accessory cell stimulated PTK activation may lead to antigen specific unresponsiveness, apoptosis or incomplete activation (40). Nevertheless, PTK activation is a pivotal first step in T-cell activation (41, 42). The TcR possesses a short intracytoplasmic tail which lacks apparent signal transduction capacity. The signaling pathways involve other surface molecules which associate with the TcR and which possess

cytoplasmic domains for PTKs (43). The CD3 complex is a multimeric heteroconjugate expressed together with the TcR, and CD3-z chain phosphorylation appears to be critical in mature T-cell signaling (44). p56lck can be activated by directly crosslinking CD4 or CD8 with specific mAbs, by APCs which bind to these surface molecules, or as demonstrated in our lab, by coculture of lymphocytes with intestinal epithelial cells (45, 33). Upon stimulation, lck autophosphorylation upregulates its own kinase activity which may lead to phosphorylation of a number of associated CD3 components (including CD3 z chain), as well as other substrates for PTK including downstream mediators of T-cell activation (46). p59fyn is phosphorylated upon crosslinking CD3 on T cells and T cell lines, and fyn kinase can be coimmunoprecipitated with CD3 components (47). Fyn activation may therefore be involved in TcR-mediated signaling, though the requirement for fyn activation remains debated, since fyn deficient T-cell lines appear to function normally, whereas lck deficient lines demonstrate activation defects (48). Nevertheless, fyn activation is known to be associated with phosphorylation of downstream effectors such as ZAP-70. In concordance with this model of T-cell activation, we observed multiple phosphorylated bands in AEC/ T-cell lysates in addition to bands corresponding to autophosphorylated lck and fyn. These additional bands are consistent with coprecipitates known to associate with activated lck, and may be phosphorylated by activated kinase in both peripheral lymphocytes and their murine counterparts. Several such coprecipitates have been previously identified and may serve as downstream mediators of PTK activity, such as the GAP-associated p62, the p62-related 68 kD protein, and hnRNP K (36). No attempt to identify coprecipitates was made in the present study. Conventional accessory cell stimulation of CD41 T-cells depends upon CD4 associated p56lck being brought into proximity with TcR/CD3 cytoplasmic domains (49). This can be initiated by the ability of class II/peptide complexes to bind both CD4 and TcR (50). The lack of observable stimulation of PTK in 3G4 cells by class II deficient A549 cell line is consistent with this model, and supports the concept that CD4-dependent stimulation by freshly isolated airway cells depends upon class II molecule expression. This is also consistent with the previously observed inhibition of AEC stimulated mixed cell cultures by anticlass II mAbs. Class I is present on both AEC and A549 and is the known ligand for CD8 (51). Although class I molecules on airway cells may interact with CD8 on 3G8 cells or with TCR and CD8 on CD81 peripheral blood T-cells, the activation of PTK in these lymphocytes does not appear to depend upon such an interaction. This is suggested by experiments in which CD8-dependent PTK activity in 3G8 cells induced by A549 could not be blocked by anticlass I mAbs, suggesting that another epithelial ligand for CD8 may exist. Additional support for this argument is that CD8 associated p56lck was not stimulated by LA-1 thyroid epithelial cells which express class I Ags. Furthermore, antibodies directed against CD1d, a nonclassical class I molecule (3C11), which appears to play a role in intestinal epithelial


cell/T-cell interaction (52, 53), did not affect PTK activity induced by A549 or AEC. This observation may be relevant to the previously observed ability of 3C11 to block T-cell proliferation in an IEC stimulated MLR, but not in an AEC or conventional accessory cell dependent MLR (55). Both lck and fyn kinase activity were stimulated in PB T-cells by A549 cells. The cell surface molecules which induce fyn activation in this system remains unclear, since our studies show that fyn activation is not an indirect consequence of lck activation. The lack of discernible class II Ag expression by A549 cells, and the lack of a significant effect of class I, leave open the possibility that some other molecule on A549 cells mediates signaling through CD3/ TcR. If such a postulated molecule is expressed on AEC (which express class II Ags) then the opportunity would exist for both conventional and nonconventional signaling through CD3/TcR. A notable difference between airway and intestinal epithelial cell populations is the finding that AEC and conventional accessory cells induce both CD4 and CD81 T-cell proliferation, whereas normal IEC induce CD81 T-cell proliferation only. Interestingly, activation of T-lymphocytes by IEC differs from activation by conventional stimulator cells, in that stimulation appears to be independent of both class II and class I MHC, and may depend upon the expression of a putative novel CD8 ligand (31). This ligand has been partially characterized as a highly glycosylated 180 kD molecule on IEC recognized by mAbs B9 and L12 (Yio, X. Y., and L. Mayer, manuscript submitted). Isolated gp180 from IEC lysates appears to bind to and stimulate CD8-lymphocytes. We have shown airway epithelial cell-dependent T-cell proliferation is blocked by mAb L12 suggesting some function for the molecule recognized by this mAb in AEC/T-cell interaction. Furthermore, L12 inhibits AEC-dependent PTK activation. By virtue of the inhibitory effect of L12 mAb, as well as the class II-dependent aspects of AEC stimulation discussed above, airway epithelial cells appear to share features with both IEC and conventional accessory cells. However, gp180 on intestinal epithelial cells appears to bind directly with CD8 and B9 Ab blocks lck activation by directly interfering with gp180/CD8 interaction. The molecule recognized by L12 mAb on AEC is different than gp180, and importantly, the effect of this antibody to block proliferation and PTK activation does not appear to be due to directly blocking an AEC ligand for CD8. Rather, L12 may recognize another molecule on AEC which is involved in AEC/T-cell interaction, such as an accessory adhesion molecule. That the epitope for L12 is shared by gp180 suggests that these molecules may be related, though such a hypothesis requires further investigation. These contrasting features of AEC and intestinal epithelial cells may reflect differences in the requirements of these cells to serve as APC in their respective mucosal microenvironments. In contrast to the GI tract, antigen load in the lung is limited and the types of antigen may vary as well. Thus, AEC might have the capacity to function as conventional APCs when challenged with a pathogenic antigen. Acknowledgments: This work was supported in part by Public Health Service Grants AI24671, DK44156, and AI23504 (L. Mayer), and the generous support

569

of the Catherine Gaisman Foundation. The authors wish to thank Drs. Daniel Krellenstein, Paul Kirschner, Steve Keller, and Jorge Camunas, and the Cardiothoracic Surgery nursing staff for their help in procuring pulmonary resection specimens.

References 1. Konig, R., L. Y. Huang, and R. N. Germain. 1992. MHC class II interaction with CD4 mediated by a region analogous to the MHC I binding site for CD8. Nature 356:796–798. 2. Cammarota, G., A. Schierle, B. Takacs, D. M. Doran, R. Knorr, W. Bannwarth, J. Guardiola, and F. Sinigaglia. 1992. Identification of a CD4 binding site on the B2 domain of HLA-DR molecules. Nature 356:799–801. 3. Kalb, T. H., M. T. Chuang, Z. Marom, and L. Mayer. 1991. Evidence for accessory cell function by class II MHC antigen-expressing airway epithelial cells. Am. J. Respir. Cell Mol. Biol. 4:320–329. 4. Glanville, A. R., H. D. Tazelaar, J. Theodore, E. Imoto, R. V. Rouse, J. C. Baldwin, and E. D. Robin. 1989. The distribution of MHC class I and II antigens on bronchial epithelium. Am. Rev. Respir. Dis. 139:330–334. 5. Mayer, L., and R. Shlien. 1987. Evidence for function of Ia molecules on gut epithelial cells in man. J. Exp. Med. 166:1471–1483. 6. Soloperto, M., L. Allegra, M. Mezzetti, A. Fasoli, and S. Matilli. 1991. Human bronchial epithelial cells modulate activation and proliferation of T cells via LFA-3. Am. Rev. Respir. Dis. 143:A226. (Abstr.) 7. Rossi, G. A., O. Sacco, B. Balbi, S. Oddera, T. Mattioni, G. Corte, C. Ravazzoni, and L. Allegra. 1990. Human ciliated bronchial epithelial cells: expression of the HLA-DR antigens and of the HLA-DR a gene, modulation of the HLA-DR antigens by g interferon and antigen-presenting function in the mixed leukocyte reaction. Am. J. Respir. Cell Mol. Biol. 3:431– 439. 8. Mezzetti, M., M. Soloperto, A. Fasoli, and S. Mattoli. 1991. Human bronchial epithelial cells modulate CD3 and mitogen-induced DNA synthesis in T cells but function poorly as antigen-presenting cells compared to pulmonary macrophages. J. Allergy Clin. Immunol. 87:930–938. 9. Paine, R., A. Chavis, D. Gaposchkin, P. Christensen, C. H. Mody, L. A. Turka, and G. B. Toews. 1992. A factor secreted by a human pulmonary alveolar-like cell line blocks T cell proliferation between G1 and S phase. Am. J. Respir. Cell Mol. Biol. 6:658–666. 10. Mattoli, S., S. Miante, F. Calabro, M. Mezzetti, A. Fasoli, and L. Allegra. 1990. Bronchial epithelial cells exposed to isocyanates potentiate activation and proliferation of T-cells. Am. J. Physiol. 259:L320–L327. 11. Kalb, T. H., F. J. Siden, Z. M. Marom, and L. Mayer. 1991. Detection of IL-6 mRNA in human airway epithelial cells. Am. Rev. Respir. Dis. 143(Suppl.):A613. (Abstr.) 12. Bonfield, T. L., M. W. Konstan, P. Burfeind, J. R. Panuska, J. B. Hillard, and M. Berger. 1995. Normal bronchial epithelial cells constitutively produce the anti-inflammatory cytokine interleukin-10 which is downregulated in cystic fibrosis. Am. J. Respir. Cell Mol. Biol. 13:257–261. 13. Smith, S. M., D. K. P. Lee, J. Lacy, and D. L. Coleman. 1990. Rat tracheal epithelial cells produce granulocyte/macrophage colony-stimulating factor. Am. J. Respir. Cell Mol. Biol. 2:59–68. 14. Standiford, T. J., S. L. Kunkel, M. A. Basha, S. W. Chensue, J. P. Lynch, III, G. B. Toews, J. Westwick, and R. M. Strieter. 1990. Interleukin-8 gene expression by a pulmonary epithelial cell line—a model for cytokine networks in the lung. J. Clin. Invest. 86:1945–1953. 15. Bellini, A., H. Yoshimura, E. Vitori, M. Marini, and S. Mattoli. 1993. Bronchial epithelial cells of patients with asthma release chemoattractant factors for T lymphocytes. J. Allergy Clin. Immunol. 92:412–424. 16. Smith, R. E., R. M. Strieter, S. H. Phan, N. W. Lukacs, G. B. Huffnagle, C.A. Wilke, M. D. Burdick, P. Lincoln, H. Evanoff, and S. L. Kunkel. 1994. Production and function of murine marcrophage inflammatory protein-1a bleomycin-induced lung injury. J. Immunol. 153:4704–4712. 17. Cromwell, O., Q. Hamid, C. J. Corrigan, J. Barkans, Q. Meng, P. D. Collins, and A. B. Kay. 1992. Expression and generation of interleukin 8, IL-6, and GM-CSF by bronchial epithelial cells and enhancement by IL-1B and TNFa. Immunology 77:330–337. 18. Schon-Hegrad, M. A., J. Oliver, P. G. McMenamin, and P. G. Holt. 1991. Studies on the density, distribution, and surface phenotype of intraepithelial class II major histocompatibility complex antigen (Ia)-bearing dendritic cells (DC) in the conducting airways. J. Exp. Med. 173:1345–1356. 19. Fournier, M., F. Lebargy, F. Le Roy Ladurie, E. Lenormand, and R. Pariente. 1989. Intraepithelial T-lymphocyte subsets in the airways of normal subjects and of patients with chronic bronchitis. Am. Rev. Respir. Dis. 140: 737–742. 20. Qian, D., I. Griswold-Prenner, M. R. Rosner, and F. W. Fitch. 1993. Multiple components of the T cell antigen receptor complex become tyrosinephosphorylated upon activation. J. Biol. Chem. 268:4488–4493. 21. Trevillyan, J. M., Y. Lu, D. Atluru, C. A. Phillips, and J. M. Bjorndahl. 1990. Differential inhibition of T cell receptor signal transduction and early activation events by a selective inhibitor of protein tyrosine kinase. J. Immunol. 145:3223–3230. 22. Veillette, A., A. Bookman, E. M. Horak, L. E. Samelson, and J. B. Bolen. 1989. Signal transduction through the CD4 receptor involves the activation

570

23. 24. 25.

26. 27. 28. 29. 30.

31. 32. 33. 34. 35.

36.

37.


of the internal membrane protein tyrosine kinase p56lck. Nature 338:257– 259. Luo, K., and B. M. Sefton. 1990. Crosslinking of the T cell surface molecules CD4 and CD8 stimulates phosphorylation of the lck tyrosine protein kinase at the autophosphorylation site. Mol. Cell Biol. 10:5305–5313. Chalupny, N. J., J. A. Ledbetter, and P. Kavathas. 1991. Association of CD8 with p56lck is required for early T cell signalling events. EMBO J. 10:1201– 1207. Chan, A. C., B. A. Irving, J. D. Fraser, and A. Weiss. 1991. The TCR zeta chain associates with a tyrosine kinase and upon TCR stimulation associates with ZAP-70, a 70Mr tyrosine phosphoprotein. Proc. Nat. Acad. Sci. USA 88:9166–9170. Sleckman, B. P., A. Peterson, W. K. Jones, J. A. Foran, J. L. Greenstein, B. Seed, and S. J. Burakoff. 1987. Expression and function of CD4 in a murine T cell hybridoma. Nature 328(6128):351–353. Ratnofsky, S. E., A. Peterson, J. L. Greenstein, and S. J. Burakoff. 1987. Expression and function of CD8 in a murine T cell hybridoma. J. Exp. Med. 166:1747–1757. Wu, R., J. Yankaskas, E. Cheng, M. R. Knowles, and R. Boucher. 1985. Growth and differentiation of human nasal epithelial cells in culture. Am. Rev. Respir. Dis. 132:311–320. Mayer, L., D. N. Posnett, and H. G. Kunkel. 1985. Human malignant T cell capable of inducing an immunoglobin class switch. J. Exp. Med. 161:134– 138. Mayer, L., E. Siden, S. Becker, and D. Eisenhardt. 1990. Antigen handling in the intestine mediated by normal enterocytes. In Advances in Mucosal Immunology. T. T. MacDonald, S. J. Challacombe, P. W. Bland, C. R. Stokes, R. V. Heatley, and A. M. Mowat, editors. Kluwer Academic Publishers, Norwell, MA. 23–28. Akiyama, T., J. Ishida, S. Nakagawa, H. Ogawara, S. Watanabe, N. Itoh, M. Shibuya, and Y. Fukami. 1987. Genestein, a specific inhibitor of tyrosine specific protein kinases. J. Biol. Chem. 262:5592–5595. 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. 178:1115–1119. Li, Y., X. Y. Yio, and L. Mayer. 1995. Human intestinal cell-induced CD81 T cell activation is mediated through CD8 and the activation of CD8-associated p56lck. J. Exp. Med. 182:1079–1088. Kefalas, P., T. R. P. Brown, and P. M. Brickell. 1995. Signalling by the p60csrc family of protein-tyrosine kinases. Int. J. Biochem. Cell. Biol. 27:551– 563. Deans, J. P., L. Kalt, J. A. Ledbetter, G. L. Schieven, J. B. Bolen, and P. Johnson. 1995. Association of 75/80 kDa phosphoproteins and the tyrosine kinases Lyn, Fyn, and Lck with the B cell molecule CD20. J. Biol. Chem. 270:22632–22638. Weng, Z., S. M. Thomas, R. J. Rickles, J. A. Taylor, A. W. Brauer, C. Seidel-Dugan, W. M. Michael, G. Dreyfuss, and J. S. Brugge. 1994. Identification of Src, Fyn, and Lyn SH3-binding proteins: implication for a function of SH3 domains. Mol. Cell Biol. 14:4509–4521. Bell, G. M., J. B. Bolen, and J. B. Imboden. 1993. Association of Src-like protein tyrosine kinases with the CD2 cell surface molecule in rat T lym-

phocytes and natural killer cells. Mol. Cell Biol. 12:5548–5554. 38. Hatakeyaamaa, M., T. Kono, N. Kobayashi, A. Kawahaara, S. D. Levin, R. M. Perlmutter, and T. Taniguchi. 1991. Interaction of the IL-2 receptor with the src-family kinase p56lck: identification of a novel intermolecular association. Science (Washington, D.C.) 252:1523–1528. 39. Juan, M., O. Vinas, M. R. Pino-Otin, L. Places, E. Martinez-Caceres, J. J. Barcelo, A. Miralles, R. Vilella, M. A. de la Fuente, J. Vives, J. Yague, and A. Gaya. 1994. CD50 (ICAM-3) stimulation induces calcium mobilization and tyrosine phosphorylation through p59fyn and p56lck in Jurkat T cell line. J. Exp. Med. 179:1747–1756. 40. Boehme, S. A., L. Zheng, and M. J. Lenardo. 1995. Analysis of the CD4 coreceptor and activation-induced costimulatory molecules in antigenmediated mature T lymphocyte death. L. Immunol. 155:1703–1712. 41. June, C. H., M. C. Fletcher, J. A. Ledbetter, G. L. Schieven, J. N. Siegel, A. F. Phillips, and L. E. Samelson. 1990. Inhibition of tyrosine phosphorylation prevents T cell receptor mediated signal transduction. Proc. Natl. Acad. Sci. USA 87:7722–7726. 42. Weiss, A., and D. R. Littman. 1994. Signal transduction by lymphocyte antigen receptors. Cell 76:263–274. 43. Pawson, T., and G. D. Gish. 1992. SH2 and SH3 domains: from structure to function. Cell 71:359–362. 44. Irving, B. A., and A. Weiss. 1991. The cytoplasmic domain of the T cell receptor zeta chain is sufficient to couple to receptor-associated signal transduction pathways. Cell 64:891–901. 45. Weiss, A. 1993. T cell antigen receptor signal transduction: a tale of tails and cytoplasmic protein-tyrosine kinases. Cell 73:209–212. 46. Chan, A. C., M. Iwashima, C. W. Turck, and A. Weiss. 1992. ZAP-70: a 70 kd protein-tyrosine kinase that associates with the TCR zeta chain. Cell 71:649–662. 47. Stein, P. L., H. M. Lee, S. Rich, and P. Soriano. 1992. p59fyn mutant mice display differential signaling in thymocytes and peripheral T cells. Cell 70: 741–750. 48. Straus, D. B., and A. Weiss. 1992. Genetic evidence for the involvement of the lck tyrosine kinase in signal transduction through the T cell antigen receptor. Cell 70:585–593. 49. Sancho, J., J. A. Ledbetter, M.-S. Choi, S. B. Kanner, J. P. Deans, and C. Terhorst. 1992. CD3-zeta surface expression is required for CD4-p56lck mediated up-regulation of T cell antigen receptor CD3 signalling in T cells. J. Biol. Chem. 267:7871–7879. 50. Miceli, M. C., and J. R. Parnes. 1991. The roles of CD4 and CD8 in T cell activation. Sem. Immunol. 3:133–141. 51. Rosenstein, Y., S. Ratnofsky, S. J. Burakoff, and S. H. Herrmann. 1989. Direct evidence for binding of CD8 to HLA class I antigens. J. Exp. Med. 169:149–160. 52. Bleicher, P. A., S. P. Balk, S. P. Hagen, R. S. Blumberg, T. J. Foltte, and C. Terhorst. 1990. Expression of murine CD1 on gastrointestinal epithelium. Science 250:679–681. 53. 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. 178:1115–1119.