Friendly rivalry

Move over kinases — a new way of phosphorylating has been found. It comes courtesy of inositol pyrophosphates and is non-enzymatic, but its existence might be highly significant in terms of intracellular signalling. Solomon Snyder and his group report their evidence in Science.

Inositol pyrophosphates, such as diphosphoinositol pentakisphosphate (5PP-InsP₅, or InsP₇) and bis-diphosphoinositol tetrakisphosphate ([PP]₂-InsP₄, or InsP₈), are formed by inositol hexakisphosphate (InsP₆) kinases (InsP6Ks). The relatively high values for the free energy of hydrolysis of pyrophosphate bonds implied that inositol pyrophosphates might be able to donate phosphates.

To test this, the authors created an artificial InsP₇ — 5β[³²P]InsP₇ — which was radioactively labelled at the β position of the pyrophosphate group. In the presence of 5β[³²P]InsP₇, several proteins from mouse and fly tissues, and from yeast — but not bacteria — were phosphorylated. Such phosphorylation was enhanced by certain phosphatase inhibitors, consistent with the normal dephosphorylation of phosphorylated proteins by phosphatases; and the process required divalent cations, preferably Mg²⁺.

So which proteins get phosphorylated by 5β[³²P]InsP₇? NSR1, a nucleolar protein that is involved in ribosome assembly and export, and YGR130c, the function of which is unknown, were both phosphorylated by 5β[³²P]InsP₇ in yeast. Both contain stretches of serines that are surrounded by acidic residues. Homology searching uncovered another nucleolar protein, SRP40, that has an acidic Ser region, and this, too, was phosphorylated by 5β[³²P]InsP₇.

Studies in kinase-mutant yeast cells ascertained that protein kinases weren't necessary for 5β[³²P]InsP₇-mediated phosphorylation. Heating increased the rate of the phosphorylation, which also implied that the process was non-enzymatic (although an initial physiological temperature was needed to ensure the phosphorylation site was in the appropriate conformation). The authors suspected that phosphorylation occurred on serine residues, as a phosphothreonine antibody failed to label 5β[³²P]InsP₇-phosphorylated proteins. They used the mammalian homologue of SRP40, Nopp140, for further analysis, and found by mutational analysis that multiple serines were phosphorylated. Acidic residues surrounding these serines, along with lysines, were also required.

Finally, Snyder and colleagues investigated the situation in intact cells. In yeast extracts that lacked InsP6K, NSR1 phosphorylation in vivo was reduced by ∼60%, which indicates that endogenous InsP₇ physiologically phosphorylates NSR1 in vivo. The inositol pyrophosphate InsP₇ is therefore a physiological phosphate donor in eukaryotic cells, which not only increases the modes of inositide-mediated signalling, but also introduces some friendly rivalry with kinases when it comes to phosphorylation.