Intestinal Intraepithelial Lymphocytes Bind to Colon Cancer Cells by HML-1 and CDlla1

Human intraepithelial lymphocytes i ILI i. predominantly CD8 ' T-lymphocytes located between intestinal epithelial cells (EC), may represent the first-line immune defense against colon cancer. The mechanism by which III bind to the colon cancer line, I) 1.1)-1. was evaluated. A larger fraction of IEL than peripheral blood mononuclear cells bound to DLD-1 monolayers (25 ±16 versus 8 ±4% binding, P < 0.05). Binding increased when DLD-1 monolayers were incubated with interferon-y but not with tumor necrosis factor-a. Similar numbers of III adhered to EC tumors, HT-29 and 5637, and the non-EC tumor, A375, but fewer bound to nonmalignant smooth muscle (HISM) and fibroblast (KD) lines (/' < 0.01). Binding of IEL to DLD-1 was reduced by monoclonal antibodies to HML-1 and CDlla (47 ±9 and 26 ±13% inhibition, respectively) and was completely eliminated by both combined (93 ±4% inhibition). YntiHML-1 also inhibited the binding of IEL to other EC tumors but did not affect binding to non-EC tumors or fibroblasLs. To conclude, the binding of IEL to EC tumors is mediated by HML-1 and CDlla [A. I. Roberts, S. M. O'Connell, and E. C. Ebert. Binding of intraepithelial lymphocytes to colon cancer cells is mediated by HML-1 and LFA-1 (abstract). Gastroenterology. 102: A685, 1992]. INTRODUCTION The early host immune responses to dysplastic intestinal EC3 and colonie adenocarcinomas are unknown. The first components of the immune system to encounter such abnormal cells are the IEL. pre dominantly CDS ' T-cells dispersed between EC. Several lines of evidence suggest that this compartment of lymphocytes may play an important role in the destruction of colon cancer: (a) the CDS4 T-cell subset of human IEL displays spontaneous cytotoxicity directed se lectively against EC tumors but not against K-562 cells, whereas the natural killer cells in peripheral blood lyse both tumor types (1); (b) colon tumor-infiltrating lymphocytes are predominantly positive for HML-1. a surface marker found on intraepithelial. but not circulating, lymphocytes (2): and (r) anti-HML-1 antibody inhibits lymphokineactivated killer activity by IEL against colon cancer target cells but not K-562 cells (3). Direct contact of lymphocytes with colon cancer cells is required for most cell-mediated antitumor immune events. Several antigenindependent binding mechanisms have been described in other sys tems. One is the attachment of the CDlla molecule of circulating lymphocytes to CD54 on epidermal keratinocytes (4), vascular endothelial cells (5). and colonie adenocarcinoma cell lines (6). Certain lymphokines. such as IFN--y, TNF-a, or interleukin 1. enhance this interaction. Another is the binding of the CD2 structure on thymocytes to CD58 on thymic EC. which results in the activation of the thy mocytes (7). IEL, too. are activated through the CD2 receptor (8). Received 8/19/92; accepted 1/21/93. The costs of publication of this article were defrayed in pan by the paymenl of page charges. This article must therefore be hereby marked ail\'erti\enient in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This project was funded by a grant from the N1H (DK 42166I. 2 To whom requests for reprints should be addressed, at Department of Medicine. University of Medicine & Dentistry of New Jersey. Rohen Wood Johnson Medical School. One Robert Wood Johnson Place-CN 19. New Brunswick. NJ 08903. 1The abbreviations used are: EC. epithelial cellls); PBMC. peripheral blood mononuclear cell(s); 1FN. Interferon; TNF. tumor necrosis factor; IEL. intraepithelial lymphocyte(s); VLA, very late activation antigen. In this study, we examined the mechanism by which IEL bind to colon cancer cells. Both pathways described above were evaluated in detail. MATERIALS AND METHODS Isolation of Lymphocytes. PBMC were separated from freshly drawn whole blood using a Ficoll density gradient (Organon Teknika. Durham. NC). IEL were separated from jejunal mucosa obtained from healthy individuals undergoing gastric bypass operations for morbid obesity. In brief, the minced mucosa was treated for 25 min at 37°Cwith 1 HIM dithiothreitol (Sigma Chemical Co., St. Louis, MO), followed by three 45-min incubations in a shaking water bath with 0.75 rmi EDTA (Sigma) in calciumand magnesiumfree Hanks' buffered salt solution (GIBCO. Grand Island, NY). The cells in the supernatams were collected and kept at 4°Covernight. IEL were separated by a Percoli (Pharmacia. Piscataway. NJ) density gradient as described previously (1.8). with the exception that cells were resuspended in the 40% Percoli layer, not the lOO'/f layer, before centrifugation. Purified IBL from above the 60% Percoli layer contained >90% lymphocytes that were 94 ±59c CD2* and 89 ±2% CDS*, consistent with the phenotype of IEL shown by immunohistochemistry (9). Magnetic Sorting. For some experiments. PBMC were enriched for CDS ' T-cells by negative selection using magnetic sorting. PBMC (10 x IO6) in O.I ml of complete medium were incubated on ice for 30 min with 5 (jg of azide-free monoclonal antibody to CD4 (AMAC. Inc.. Westbrook. ME) and then washed three times. Cells were resuspended to 4 x 106/ml. and 0.5 ml of prewashed magnetic goat anti-mouse IgG suspension (Collaborative Biomed icai Products. Bedford. MA) was added. After a 20-min incubation on ice. magnetic particle-cell complexes were removed by application of an Alnico V magnet to the side of the tube for 10 min at 4°C.After separation. PBMC preparations were < 17c CD4 * by immunofluorescence. Binding Assay. Various cell lines (American Type Culture Collection. Rockville, MD) were seeded into flat-bottom microwells in complete medium: RPMI-1640 containing 107r fetal bovine serum (Hyclone Laboratories. Logan. UT). \7c antibiotic-antimycotic solution and glutamine (GIBCO). and 10 mw 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid (Sigma). Some cultures were supplemented with IFN-y. TNF-a. or both lymphokines combined (each at 100 units/ml: Amgen. Thousand Oaks. CA). All cultures formed confluent monolayers after an overnight incubation. The supernatant was aspirated, and 4 x IO5 lymphocytes were labeled with [MCr]sodium enrÃ3mate (New England Nuclear, Boston. MA) were added to each well. After incubating 2 h at 37°C, nonadherent cells were removed by gentle washes accomplished by repeatedly adding and decanting warm complete medium. A 27c cetrimide solution (Sig ma) was then added to lyse the adherent cells, the contents of each well were collected, and the radioactivity was counted. After the spontaneous release of radiolabel was adjusted for. the fraction of lymphocytes bound to the monolayer was calculated as a percentage of the total added. In some experiments, monoclonal antibodies to the following surface mark ers were added to the cell cultures 10 min before adding labeled lymphocytes and left in for the duration of the assay: CD2 (Til; Coulter Immunology, Hialeah. FL); CDlla (LFA-1). CD54 (ICAM1), CD58 (LFA-3), HML-1 (AMAC. Inc.). and VLA-1 (TCell Sciences. Cambridge. MA), with or without a 300-fold molar excess of human IgG-Fc fragment (Accurate Chemical. Westbury. NY). Flow Cytometrics Analysis. Lymphocytes were stained by indirect im munofluorescence with various monoclonal IgG antibodies, followed by fluorescein-conjugated goat anti-mouse IgG (Coulter). Fluorescence was ana lyzed using a Coulter Profile analytical flow cytometer with a 25-mW argon laser. Cell size was measured by forward angle light scatter, and inlracellular granularity differences were measured by right angle light scatter. These pa rameters were used to identify and gate the major cell subpopulations for