What does a tiny aquatic flowering plant have in common with a process for repairing broken DNA molecules? The answer is that both require a phosphorylated cyclic alcohol called inositol hexakisphosphate (InsP6), and, for the DNA-repair process at least, a report by Hanakahi and colleagues in the 15 September issue of Cell explains why.

The story revolves around the catalytic subunit of the DNA-dependent protein kinase(DNA-PKcs), which helps to repair double-stranded DNA breaks using the non-homologous end-joining (NHEJ) pathway. The DNA-PKcs is first targeted to each of the severed ends by a complex of the Ku70 and Ku80 proteins (see figure). DNA-PKcs then recruits a complex of DNA ligase IV and XRCC4, resulting in rejoining of the broken ends.

In vitro assays have shown that although all of these players — Ku70, Ku80, DNA-PKcs, XRCC4 and DNA ligase IV — are needed for NHEJ, they are not sufficient. So something is missing, and to discover what, Hanakahi et al. used an in vitro complementation assay. They took cell-free extracts that could promote NHEJ, fractionated them on a phosphocellulose column, then added back various combinations of the fractions to restore the NHEJ activity. In so doing, the authors identified a fraction containing the elusive factor, which they named stimulatory factor A (SFA).

The task of identifying SFA involved eight purification steps followed by a barrage of tests, including protease digestion, nuclease digestion and even boiling. But the activity survived all of these insults intact, indicating — unexpectedly — that SFA is neither a protein nor a nucleic acid.

So what is it? The authors concluded that the active ingredient is, in fact, InsP6, and showed that the component of the NHEJ machinery to which it binds is probably DNA-PKcs. They suggest several possible functions for InsP6 in NHEJ, including the idea that it converts DNA-PKcs from an inactive to an active form.

Interestingly, DNA-PKcs, as well as several other DNA-repair proteins such as ataxia telangiectasia mutated (ATM) and ATR, contains motifs characteristic of phosphatidylinositol-3-OH kinases (PI(3)K). As none of these proteins has lipid kinase activity, why have they retained such motifs? Hanakahi et al. speculate that such proteins may have evolved from a common ancestor with both protein and lipid kinase functions, and that mutation of the PI(3)K domain caused these proteins to lose their lipid kinase activity but retain the ability to bind headgroups containing inositol polyphosphates. However, there is no evidence that this is where the InsP6 binds DNA-PKcs — indeed, there is a big charge difference between the phosphatidylinositol-3,4,5-trisphosphate headgroup and InsP6 — and it may turn out that InsP6 binds to an allosteric activation site on DNA-PKcs.

Either way, InsP6 is a molecule of the moment. As well as making up an enormous 60% of the volume of the aquatic plant duckweed, where it acts as a kind of ballast, InsP6 is thought to act as an antioxidant and phosphate storage source in plant seeds. In mammalian cells it has been implicated in regulating growth, inflammation and nervous transmission, and it is even thought to be involved in export of messenger RNA from the nucleus. Nonetheless, the link to DNA repair comes as a surprise, and should trigger a search for InsP6 -binding sites in other members of the PI(3)K family.