Reporting in Science, Levine and colleagues provide new insights into how cells coordinate changes in the composition of their cellular membranes with lipid metabolism. In yeast, in response to inositol (a precursor of the phospholipid phosphatidylinositol), it is known that the endoplasmic reticulum (ER)-bound transcription factor Opi1 becomes active and represses the transcription of a gene, INO1 . INO1 encodes an enzyme that is important for inositol synthesis. But, to induce this negative-feedback loop, what signal does inositol generate and how does Opi1 sense this signal?

The authors first showed that inositol caused green-fluorescent-protein–Opi1 to translocate from its ER location to the nucleus. As the timing of this change correlated with a decrease in INO1 mRNA levels, they proposed that this change is physiologically relevant. But what is Opi1 sensing?

By labelling the phospholipid pool in the ER with 32P-orthophosphate, Levine and co-workers showed that the nuclear translocation of Opi1 was temporally related to large changes in the levels of phosphatidylinositol and its precursors — the latter were consumed to produce the former. So, they wondered whether Opi1 might be directly sensing these changes.

It has long been thought that phosphatidic acid (PA), a precursor of phosphatidylinositol, is involved in the regulation of lipid metabolism, so the authors tested whether Opi1 could bind PA. They showed that Opi1 could indeed bind PA, both directly and specifically, and that Opi1 could detect PA-rich membranes in vivo.

Except for the presence of one or more basic residues, there is no known PA-binding motif. However, Levine and colleagues showed that Opi1 seems to contain two PA-binding sites — one in its carboxy-terminal half and another in a basic domain in the second quarter of the protein. When they mutated this basic domain to disrupt the PA-binding ability of Opi1, they found that Opi1 was activated and translocated to the nucleus in the absence of inositol.

So, this work has clarified the details of a negative-feedback loop that controls phospholipid metabolism in yeast. The ER pool of PA directly binds to Opi1 to maintain it in an inactive state, but in the presence of inositol, this pool is consumed. The decreasing PA levels result in the release of Opi1 from the ER and its translocation to the nucleus, where it can repress the synthesis of inositol. Although many proteins are known to bind PA, the physiological significance of this has been unclear. This study, which has shown a physiological response to changing PA levels, therefore indicates that specific pools of PA might have important signalling roles in other cells.