The sea squirt in question, Botryllus schlosseri, exists as a flower-shaped accumulation of petals, or zooids. When two clusters of zooids meet one another they either fuse or reject one another. For fusion to occur, the two must share at least one allele of the Fu/HC locus.
The molecular identity of Fu/HC remains unknown, and the search for direct relatives of MHC proteins, which determine transplant rejection in humans, has not produced any candidates. The recent genome sequencing of the sea squirt Ciona intestinalis also failed to turn up any MHC proteins, although Ciona is solitary and thus lacks the rejection reaction of Botryllus.
Khalturin and Bosch searched for genes whose expression changed during the rejection process. One of the down-regulated genes encodes BsCD94–1, a protein on the surface of Botryllus blood cells that is very similar to the vertebrate NK cell receptor CD94. The group does not yet have functional evidence tying the protein to rejection, but Bosch says “the structural similarity is absolutely convincing.”
The pattern of recognition in the two systems is also a match. In Botryllus, an A/B genotype rejects C/D but fuses with B/C. Likewise, NK cells kill only if CD94 fails to recognize self (in the form of class I MHC), whereas the B and T cells of the vertebrate adaptive immune system attack as long as they see any trace of nonself.
Adaptive immunity has been traced back only as far as jawed vertebrates. The new work suggests that a type of NK cell existed much earlier, but it may have been specialized not for immunity but for adjudicating land grabs.
And even those land grabs have a twist. Others have found that, although one of the colonies often dies off after fusion, its germ cells can parasitize the survivor. Hence, the sea squirt chooses to fuse only with genetic relatives, so that after fusion its energy is devoted to propagating infiltrating germ cells that are at least related. Natural killing, it seems, began with family loyalty. ▪