Study of the humoral and cellular immune response : to Saccharomyces cerevisiae in man
Saccharomyces cerevisiae (bakers'/brewers' yeast) is a ubiquitous dietary constituent in the developed world. Previous studies, using semi-quantitative ELISA techniques, suggested that patients with Crohn's disease have higher titres of IgG and IgA isotypespecific antibodies to this yeast than are found in normal control subjects or patients with ulcerative colitis. For this study, in order to allow more stringent assay standardisation and more meaningful numerical comparison of the relative antigen-binding capacities of different sera, a quantitative ELISA was developed for measurement of anti-yeast antibodies, using a soluble extract of yeast (sacc) as the antigen. The finding of raised levels of yeast antibodies in Crohn's disease was confirmed, and the data suggest that this may be related to the presence of disease in the small bowel, although this latter observation did not reach statistical significance. Patients with chronic liver disease also had higher antibody levels than controls, but less markedly so than in Crohn's disease. When sera were tested in a similar assay for antibodies to bovine casein, no difference was found between controls and the Crohn's or liver disease group. The response of peripheral blood mononuclear cells (PBMC) to sacc was examined using a proliferation assay measuring uptake of tritiated thymidine. Cells from normal controls demonstrated dose-dependent proliferation, the time-course of which resembled that obtained with known recall antigens. Following separation of cell populations by rosetting with sheep erythrocytes, the responding cells were shown to be T-lymphocytes and the magnitude of the response was sensitive to the number of antigen-presenting cells present in the culture. When positive selection with immunomagnetic beads was used to further separate T-cells into highly purified CD4+ and CD8+ populations, responsiveness to yeast co-separated with the CD4+ subset. Following negative selection of cells expressing CD45RO or CD45RA, responsiveness was largely, but not exclusively, confined to the CD45RO+ population. Limiting dilution analysis of peripheral blood T-cells gave estimates of the sacc-specific precursor cell frequency in keeping with values previously reported for recall antigens, although the experimental data could not be shown to conform to single-hit kinetics. By sequential stimulation in long term culture, it was possible to obtain populations of cells which were uniquely responsive to sacc but unresponsive to other recall antigens. At some concentrations of sacc, proliferation responses of PBMC from Crohn's disease patients were higher than those in normal subjects, but the difference was not convincing overall. Digestion of sacc with pronase abolished the T-cell response but left specific antibody-binding intact, supporting the suggestion that antibody recognition is dependent on carbohydrate epitopes. Yeast cell wall mannan is implicated as the likely site of B-cell epitopes; evidence pertaining to T-cell epitopes is less conclusive. Thus, this study provides evidence that immune sensitisation to a common dietary constituent frequently occurs in the normal population, leading to detectable humoral and cellular immune responses. The T-cell response appears to be genuinely antigen-specific, and not due to non-specific lymphocyte activation. The gastrointestinal lymphoid system may be the site at which primary sensitisation occurs. In patients with Crohn's disease, the humoral response is enhanced, possibly as a consequence of inflammatory processes in the small bowel.