The above new findings concerning the immunological mechanisms governing mucosa, immune responses and oral tolerance in TCR-transgenic mice, as well as those operative in mice with experimental colitis, greatly expand our understanding of the processes that normally control mucosal inflammation and possibly other types of inflammation as well (Fig. 1). They indicate that, in the nondiseased mouse, ingested proteins evoke a Th1-cell (IFN-gamma) response in the mucosal follicles that is quickly counter-regulated by induction of T-cell anergy/deletion, if this Th1-cell response is inhibited (experimentally by anti-IL-12), TGF beta-producing cells appear, and these are capable of active immune suppression. This reciprocal relationship between IFN-gamma production and TGF-beta production is further supported in mouse models of mucosal inflammation. Thus, in the TNBS-colitis model, there is direct stimulation of the immune cells in the lamina propria as a result of diffuse haptenization of mucosal proteins, which leads to a massive Th1-cell response capable of overwhelming any suppressive counter-regulatory mechanisms normally generated in the PPs. This highly polarized Th1-cell response is controlled only by direct abrogation of IL-12 production with exogenous administration of anti-IL-12, or with indirect inhibition of this response via induction of oral tolerance and accompanying production of TGF-beta (Refs 6-8). The data obtained from this model are consistent with those obtained with another model of intestinal inflammation--inflammation in severe combined immunodeficiency (SCID) mice following lymphoid repletion with CD45Rbhi (naive) T cells. In this model, inflammation is again mediated by Th1 cells and is prevented by co-repletion with CD45Rbhi (memory) T cells, which appear to work by secreting TGF-beta (Refs 9, 10). Thus, a common feature of the various experimental models of intestinal inflammation studied to date is the Yin-Yang relationship of IFN-gamma and TGF-beta, with the former being proinflammatory and the latter anti-inflammatory. Is the IFN-gamma TGF-beta dichotomy that is evident both in the normal state and in models of inflammation simply a reflection of an underlying Th1 Th2 dichotomy? The answer to this important question is not yet known. Thus, while it is clear from the in vitro studies already discussed that IL-12 and/or IFN-gamma inhibit TGF-beta production, the role of IL-4 in this process is more elusive. These in vitro studies indicate that IL-4 is not required for TGF-beta production, a finding that is consistent with studies in which the transfer of CD45Rbhi, T cells from IL-4-/- mice protected SCID mice from colitis induced by CD45Rbhi T cells. However, the addition of IL-4 to in vitro cultures containing anti-IL-12 augmented TGF-beta production, most probably by IL-4 acting as a growth factor for TGF-beta-producing cells rather than as an inducing factor (T. Marth et al., unpublished). Obviously, more work will be necessary to resolve this issue. Finally, it should be noted that the above considerations apply to human inflammatory diseases of the gastrointestinal tract, such as Crohn's disease. Recently, it has been shown that T cells extracted from Crohn's disease tissues manifest skewed Th1-cell responses. The hypothesis can therefore be put forward that this disease results from a dysregulated Th1-cell response to ubiquitous mucosal antigens that is not appropriately controlled by normal counter-regulatory mechanisms. Interventions that artificially bring the excessive Th1-cell response back into balance, such as administration of IL-12 antagonists, should therefore find a central place in the treatment of the disease.