Since the embryo and its associated extra-embryonic tissue express genes that are derived from both maternal and paternal alleles, and since this conceptus embeds within maternal uterine endometrium as a "graft" for the whole of the rest of pregnancy, it is unsurprising that much research into the "fetal allograft" problem [5] has been undertaken by reproductive immunologists over the last 40 years. Remarkably a quite unexpected finding that may resolve this question was published in 1998. Munn and colleagues [9] showed (in mice) that a tryptophan-catabolizing enzyme indoleamine 2,3-dioxygenase [14] was required to be functioning presumably in the placenta if the expected immune response (which they showed would otherwise necessarily result in fetal rejection) was to be averted. Table 1 indicates their key experimental finding.The strength of the experimental design comes from the availability of both a specific pharmacological inhibitor (1-methyl-tryptophan) [2]of the indoleamine 2,3-dioxygenase enzyme and "smart" transgenic animals that provide unambiguous controls. Since for allogeneic matings (but not for syngeneic ones, where no immunological problems arise) pregnancies were successful only in animals with normally functioning placental indoleamine 2,3-dioxygenase, this seminal paper did not resolve how tryptophan metabolism is responsible for averting the immune response to pregnancy. Subsequently Munn et al. [8] provided strong evidence that indoleamine 2,3-dioxygenase, simply through localized tryptophan depletion, is able to suppress T lymphocyte proliferation and hence produce "immunosuppression by starvation" . Figure 1A shows that indeed the ability of human T cells to respond to a stimulus such as that produced by the nonspecific mitogen PHA is uniquely and absolutely dependent on the presence of (low concentrations of) tryptophan. It is not yet clear how this tryptophan-dependent restriction of T cell proliferation is brought about, but it is without doubt a novel and important aspect of T cell biology. However, of general interest here is the physiological context in which this key enzyme, indoleamine 2,3-dioxygenase, works: how can the placenta play this central role in regulating the maternal immune response mounted against it while at the same time allowing normal fetal nutrition?In pregnant women indoleamine 2,3-dioxygenase is expressed in the placenta particularly in the syncytiotrophoblast, the fetal-derived component that is in direct contact with the maternal endometrium during early pregnancy (including implantation) and subsequently later in pregnancy with the maternal circulation flowing through the intervillus space. (Although increased maternal tryptophan catabolism and production of the indoleamine 2,3-dioxygenase product kyneurine was observed by Schrocksnadel et al. [11], unsurprisingly these investigators did not at that time realize the possible significance of these findings.) However, the enzyme indoleamine 2,3-dioxygenase is cytoplasmic, which immediately raises two interesting...