Many species go to great lengths to prevent or at least minimize mating between close relatives — or, in the case of plants, with themselves. Nematodes, on the other hand, show a blatant disregard for this taboo, as many species exist largely as self-fertile hermaphrodites. Their existence, however, and our detailed genetic knowledge of one such worm, Caenorhabditis elegans, is a boon to those studying sex determination. But is it possible to describe a conserved pathway for hermaphroditism? By comparing the sex determination pathways of two sister species of Caenorhabditis, Sudhir Nayak and colleagues provide functional evidence that this is not the case, and that hermaphroditism has probably evolved by different genetic routes in the two species.

Thirty-one genes are known to be involved in establishing the germline and somatic sex of C. elegans. Although 30 of these genes have counterparts in the relative Caenorhabditis briggsae, the authors have found that one gene, fog-2 , has no convincing orthologue in this species. This is puzzling given the fact that FOG-2 is needed to trigger spermatogenesis in the C. elegans hermaphrodite: here, a complex between FOG-2 and the RNA-binding protein GLD-1 shuts down the synthesis of the feminizing TRA-2 protein, leading to sperm production. Given that C. briggsae hermaphrodites obviously produce sperm, and that this organism has presumed functional copies of all other known sex-determining genes, including gld-1, how does this species establish the male-germ-cell fate?

RNA interference (RNAi) of the gld-1 transcript shows that this gene has opposite functions in the two Caenorhabditis species: whereas, as expected, knocking-down gld-1 in C. elegans hermaphrodites led to a feminized germline, RNAi masculinizes the germline of C. briggsae hermaphrodites, as sperm is produced at the expense of oocytes. The difference in gld-1 function between the two worm species is apparently limited to germline sex determination, as its requirement for oogenesis and meiotic progression has been conserved. However, this work does not tell us how GLD-1 might direct sperm production in C. briggsae in the absence of FOG-2 (and whether, in fact, it does), although the testing of some hypotheses is underway.

The lack of fog-2 in C. briggsae and the opposing roles of gld-1 in germline sex determination indicate that hermaphroditism has evolved independently in these two species, and by extension in the whole Caenorhabditis clade. In fact, this work bolsters the existing view that sex determination mechanisms are incredibly labile, and depend more on a permissible anatomy and physiology (which is broadly similar across the clade) than on conserved genetic relationships.