A review of: Newberry RD, Stenson WF, Lorenz RG 1999 Cyclooxygenase-2-dependent arachidonic acid metabolites are essential modulators of the immune response to dietary antigen. Nature Med 5:900–906.

Studies of the basis of immunological tolerance are of increasing concern to pediatricians because of a reported increase of immune dysregulation in developed-world children. Apparently, this has not occurred within the developing world, except in relatively few privileged children. The effects of altered early environmental conditioning have been observed in children within former East Germany following reunification (1), and are beginning to be felt in other countries with rapid social progression, such as Singapore (2). The consequences potentially go beyond classic atopy to include marked increases of more severe immunopathology such as childhood Crohn's disease, in which tolerance to the enteric flora is lost, and possibly autism, in which there is initial evidence of dysregulated mucosal immunity.

Recently there has been much discussion as to whether a lack of early infectious challenges could subvert a system of early immune priming, established over millenia in which childhood survival was an achievement rather than a right (3). This study, in which local inflammation is seen to be required to establish oral tolerance, is thus of direct conceptual relevance to what may become a major challenge facing pediatricians in the next century.

Great advances have been achieved in understanding of the mechanisms of oral tolerance, which is essentially an actively mediated process. High-dose dietary antigen causes T cell anergy, and consequent generation of the anti-inflammatory cytokine IL-10, while low-dose antigen induces generation of TH3 cells that secrete transforming growth factor-β (TGF-β) (4). Both cell types are critical in preventing inflammation by the process of “bystander tolerance,” in which they suppress reactions of all surrounding lymphocytes (5). Remarkably, infectious challenge seems to be required to establish oral tolerance. Animals maintained germ-free might not develop tolerance adequately, although the exact mechanisms have been difficult to unravel (6, 7).

The beauty of this study lies in the design, in which transgenic mice have been engineered with highly restricted immune reactivity, possessing only CD4 αβ T cells reactive to hen egg lysozyme (i.e. unable to respond to enteric bacteria or other antigens). Feeding of hen egg lysozyme to these mice caused no apparent problems, and necropsies showed normal villous architecture. However concurrent administration of indomethacin, an inhibitor of inducible cyclooxygenase-2 (COX-2), and thus of arachidonic acid metabolism, broke tolerance and induced marked enteropathy. The authors showed that mononuclear cells (largely macrophages) from the intestinal lamina propria produced COX-2 dependent factors, in particular prostaglandin PGE2, at a level 100 times that of splenocytes on activation.

PGE2 has a marked in vitro effect of suppressing antigen-specific T cell proliferation, as well as increasing IL-10. The authors demonstrated that co-administration of a specific COX-2 inhibitor, NS-398, with hen-egg lysozyme also abrogated tolerance and led to an enteropathy. Thus the establishment of oral tolerance, and the induction of normal bystander tolerance, within the small intestine of newborns may require a major contribution from the innate immune system (PGE2 production by mucosal macrophages in response to lipopolysaccharide from luminal bacteria). In this context, the major changes that have occurred in the initial gut colonisation of developed world, but not developing world, infants within the last two generations may be of more than passing significance (8, 9).