Department of Immunology and Bacteriology, University of Glasgow, Western Infirmary, Glasgow, Scotland
Address for correspondence: Allan McI. Mowat, Department of Immunology and Bacteriology, University of Glasgow, Western Infirmary, Glasgow, Scotland. Voice: +44-141-2112728; fax: +44-141-3373217. a.m.mowat{at}clinmed.gla.ac.uk
Oral tolerance was first detailed almost 100 years ago, and
since then, it has been shown repeatedly that feeding a wide
variety of nonpathogenic antigens can inhibit subsequent systemic
immune responses. All systemic immune responses are susceptible,
but the degree and scope of the suppression depends on the nature
and dose of the fed antigen. Oral tolerance has been described
in most mammals, including humans, and it may be the homeostatic
mechanism that prevents hypersensitivity to food antigens, as
is found in celiac disease. A similar process may prevent the
aberrant immune responses to commensal bacteria that occur in
inflammatory bowel disease. The ability of oral tolerance to
modulate experimental models of autoimmune and inflammatory
disease has led to clinical trials in such diseases as rheumatoid
arthritis, multiple sclerosis, and type I diabetes, with only
variable success. Despite intense research, the exact mechanisms
responsible for the systemic tolerance and the reasons why tolerance
is the default response to many fed antigens remain controversial.
Early studies suggested that CD8
+ "suppressor" T cells were
important, but it is now accepted that it may involve either
anergy/deletion of CD4
+ T cells, or the induction of regulatory
CD4
+ T cells that produce IL-10 and/or TGFß. There
may also be a role for CD4
+ CD25
+ T
reg, but how and when all
these different mechanisms operate is still unclear. The ability
of fed antigens to induce tolerance probably reflects their
uptake by "quiescent" antigen-presenting cells in the intestine,
with presentation to specific CD4
+ T cells in the absence of
costimulation, or with the involvement of inhibitory costimulatory
molecules. Dendritic cells in the Peyer's patches or mucosal
lamina propria are the most likely APCs involved, but it remains
to be determined exactly where these interactions occur and
what the precise nature of the relevant dendritic cells is.
