Immunologists have never found it easy to explain why the semi-allogeneic fetus is not attacked in placental mammals by the maternal immune system. How is a response against paternal antigens avoided? Most researchers agree that local effects at the maternal–placental interface are important. Hobbs et al. suggest in Cell that the vertebrate-specific amino-terminal domain of the TATA-box-binding protein (TBP) regulates a placenta-specific β2-microglobulin (β2m)-dependent process that is used to evade a maternal rejection response.

The authors generated mice that have a modified Tbp allele (TbpΔN) that lacks the amino-terminal domain. More than 90% of TbpΔNN fetuses died in mid-gestation. Histopathology indicated that fetal death was the result of a placental insufficiency. Moreover, the physiology of mutant fetuses was normal, and the rare homozygous adults were healthy. This indicates that the mutation primarily affects a process that is required for placental, but not fetal, development.

Next, the authors showed that the defect could be complemented by an immunodeficient maternal environment. When the TbpΔN mutation was bred into recombination-activating gene 1 (Rag1)−/− or severe combined immunodeficient (SCID) mice, most of the TbpΔNN fetuses survived, which indicates that the maternal immune response has an important role. This was confirmed by the observation that the defect can also be complemented by knockout of β2m. As β2m is required for the assembly and surface expression of MHC class I molecules, the defect probably involves inappropriate antigen presentation.

The authors suggest that the TBP amino-terminus downregulates antigen presentation by MHC class I. They speculate that the amino-terminus might have co-evolved with the MHC system to regulate its expression. In mice, the placenta is simply the first vital location of this function. It will be interesting to determine whether mutant adults have defects in other MHC class-I-dependent responses, such as susceptibility to intracellular pathogens.