Cytokine Production by Human Colonic Intraepithelial Lymphocytes in Controls and Ulcerative Colitis

Using an ELISA technique, concentrations of γ-interferon and interleukin-2 were assayed in the supernatants of colonic intraepithelial lymphocytes cultured with or without phytohaemagglutinin (PHA). IntraepitheHal lymphocytes produced low concentrations of γ-interferon and interleukin-2 when stimulated with PHA, but significantly more than when unstimulated (p < 0.05). There was no difference in production of these cytokines by IEL from control or inflammatory bowel disease.


Introduction
Intraepithelial lymphocytes (IEL) are found interspersed between the epithelial cells of the gastrointestinal tract and so bear a close spatial relationship to luminal antigens. The precise function of IEL in the human is not known, although they may have a role in the induction of gut tolerance to luminal contents. 1,2 Cytokines are soluble factors released by immune cells, especially macrophages and lymphocytes. One hypothesis for the pathogenesis of ulcerative colitis (UC) is that abnormal release of cytokines may result in disordered immunoregulation. Murine IEL do not spontaneously release interleukin-2 (IL-2) but do so when stimulated. -5 A ,-interferon-like factor has also been shown to be produced after stimulation with mitogens.
Human small bowel IEL have been reported to produce IL-2 in response to stimulation with phytohaemagglutinin (PHA) and T-interferon (TIFN) in response to PHA and sheep red blood cell (RBC) lysates. However, no data are available concerning cytokine production by human colonic IEL. UK) without calcium and magnesium, and supplemented with gentamycin 50 t.tg/ml, penicillin 100 U/ ml and streptomycin 100 l.tg/ml (HBSS). The mucosa was dissected and following further washing the mucosal strips were incubated with RPMI supplemented with 10% foetal calf serum, antibiotics and 1 mM dithiothreitol (Sigma) for 10 min. The mucosal strips were then washed a further three times in HBSS and incubated with 0.75 mM EDTA in a shaking water bath at 37C for three periods of 30 min each at 140 oscillations per min. Following each incubation the supernatant containing the 1EL was pooled, centrifuged at 500 x g for 8 min and the pellet washed three times in RPMI. The cells were resuspended in RPMI and stored overnight at 4C. Further purification was achieved by passing this crude preparation down a glass wool column (Sigma) and performing a two-step Percoll gradient of 44% and 67.5%. The IEL were recovered from the interface.

Immunofluorescence and FACS analysis:
Immunofluorescent staining was performed as described previously. Briefly, cells were suspended in PBS/0.1% sodium azide to a concentration of 106 cells/ml. Aliquots of 100 lal were incubated with 8 of monoclonal antibody in the dark at 4C for 45 min. The cells were then washed with PBS, centrifuged and resuspended in 500 btl PBS, paraformaldehyde 1%. Flow cytometric analysis was performed using a FACScan (Becton Dickinson).
Preparation of lymphokine-containing supernatant: IEL at a concentration of 1 x 106 were cultured in l ml aliquots in 24-well plates (Linbro, Flow Laboratories, UK) with or without 10 btg/ml PHA (Wellcome Laboratories, Beckenham, UK) for 72 h at 37C and 5% CO2. The supernatants were then harvested, centrifuged and passed through a 0.22 btm filter and stored at-20C until analysis was performed.
Cytokine assays: 3,IFN was measured by an enzyme amplified sensitivity immunoassay of proven specificity (Medgenix Diagnostics, Belgium). IL-2 was measured by a sandwich enzyme immunoassay of proven specificity (
Induction of HLA-DR molecules by culture supernatant on HT29: Supernatants of unstimulated IEL when incubated with HT29 cells produced no detectable expression of HLA-DR molecules, as assessed by flow cytometry. However, the supernatants of stimulated control IEL produced 5% HLA-DR expression of HT29 (Fig. 3). This effect was also observed using supernatants of stimulated IEL which had been irradiated (2 500 Rad) prior to culture (data not shown).

Discussion
Rat IEL produce a yIFN-like factor capable of producing Ia expression on IEC 17 cells. Lymphocyte proliferation was not essential for the production of this factor. In mice, Con A stimulated IEL to produce moderate levels of IFN. Human small intestinal IEL have been reported to produce 7IFN after stimulation with PHA. 8 This production of 7IFN is increased many times by the addition of sheep RBC lysates.
The present data show that colonic IEL produce very low amounts of yIFN. Although there was a significant increase in 7IFN production following stimulation with PHA, the absolute amounts of yIFN produced were persistently low. This is not due to the failure of lectin binding. 1 The amount of yIFN produced by stimulated IEL is of sufficient quantity to induce Class II molecules on HT29 cells. Class II bearing epithelial cells can in turn cause further activation of IEL. n In contrast to a previous study on cytokine production by lamina propria lymphocytes there was no significant difference in yIFN production by IEL between control and IBD patients. 12 Most animal studies have shown that IEL do not produce IL-2. Human small bowel IEL produce similar amount of IL-2 to peripheral blood T cells following stimulation with mitogen. These data confirm the finding that human colonic IEL produce IL-2. In addition we have shown that there is no difference in IL-2 production by colonic IEL in control and UC patients.
It has recently been proposed that helper T.
(CD4/) cells can be classified with relationship to their cytokine production profile. 13  Human CD8 clones have also been analysed in a similar manner. Lepromin specific CD8 T suppressor clones from patients with lepromatous leprosy produce IL-4 and little or no yIFN (Type 2 pattern of lymphokine production). In contrast, alloreactive HLA B27-specific human CD8 CTL clones produce yIFN but no IL-4 (Type 1 lymphokine pattern). In these experiments, suppression was IL-4 dependent and not necessarily related to phenotype, as CD4 clones which secreted IL-4 also caused suppression. 14,15 Human colonic IEL have recently been shown to have potent suppressive properties on the proliferative responses of autologous LPL. 1,2 This effect was found to be CD8-dependent, 78-independent and mediated by a soluble factor. The data presented in this paper indicate that IEL produce only small quantities of yIFN and IL-2, which indicates that they could have a Type 2 pattern of lymphokine production. Were this to prove to be the case, it may explain the potent down-regulatory properties of IEL, on the basis of IL-4 and/or IL-10 secretion.
In conclusion, colonic IEL produce the cytokines yIFN and IL-2, although in small amounts. There is no significant difference in production of yIFN and IL-2 in control and UC patients.