Cell-cycle control in eggs and early embryos has its own set of rules. In Drosophila melanogaster, the first mitosis is delayed until after fertilization, after which S phase and M phase are alternated for thirteen simplified cell cycles. Maternal mutations in three genes — pan gu ( png ), plutonium ( plu ) and giant nuclei ( gnu ) — disrupt this control, yielding eggs that replicate their DNA before fertilization, and embryos with fewer, larger nuclei; but how? A paper in Development provides the first mechanistic insights into this conundrum.

Fenger and colleagues found that png encodes a serine/threonine protein kinase expressed in the early embryo. It binds PLU and the complex has kinase activity in vitro. How does this complex control the cell cycle? A clue came from examining png mutant embryos: these have lower levels of cyclins A and B, and lower CDC2 kinase activity, so they carry on synthesizing DNA but can't undergo mitosis. png, therefore, is needed to limit S phase and promote mitosis, which can be achieved by maintaining mitotic cyclin levels. However, no direct interaction between PNG–PLU and cyclins or CDC2 was found. Instead, the authors speculate, PNG–PLU might stabilize cyclins by blocking their access to the protein degradation machinery.

Many questions remain; for example, what are PNG's substrates? And does GNU interact with PNG–PLU? But one mystery has been solved: in vertebrates, the dual-function kinase Mos (see the Timeline by Yoshio Masui on page 228) prevents division of unfertilized eggs. Invertebrates don't have Mos, but we now know that PNG can carry out some of its functions in flies.