Sleep is regulated in two opposing ways: by the circadian clock and by the homeostatic sleep drive. The molecular underpinnings of the homeostatic mechanism are poorly understood. Koh et al. have now identified sleepless (SSS), a protein that is required for homeostatic sleep pressure in Drosophila melanogaster.

In order to identify the molecular factors that drive animals to sleep, the authors carried out a genetic screen to find D. melanogaster mutants that exhibit reduced daily sleep. The mutant with the lowest amount of daily sleep was named sleepless (sss), as sleep was reduced to just 15% of normal levels. In these mutants, the sss gene carried a P-element insertion (sss P1) in the coding region that resulted in the loss of SSS protein. Another insertion (sss P2) in the 3′ untranslated region of the same gene resulted in a moderate reduction of SSS protein levels and in only a minimal reduction in sleep amount. Furthermore, sss P2/sss P1 transheterozygous mutants exhibited ∼30% of normal daily sleep amount and showed greatly reduced levels of SSS protein. These findings suggest that the amount of sleep directly correlates with SSS protein levels. This hypothesis is further corroborated by the fact that sleep duration in sssP1 mutants was restored to wild-type levels when the P-element insertion was excised or when wild-type sss was reintroduced transgenically.

Further characterization of SSS showed that it is a brain-enriched, glycosylphosphatidylinositol (GPI)-anchored cell-surface protein, the levels of which do not change throughout the circadian cycle. The GPI anchor could be cleaved by phosphatidylinositol-specific phospholipase C (PLC) to release SSS from the cell surface. Other GPI-anchored proteins function as ligands or co-receptors and can also act as diffusible signals when cleaved by PLC.

Next, the authors analysed the contribution of SSS to sleep homeostasis. Whereas control flies showed substantial rebound sleep after deprivation, sssP1/sssP2 flies and sss P2 flies showed little or no rebound, providing evidence that SSS has a role in sleep homeostasis.

Interestingly, a previously described gene called quiver (qvr) has been mapped close to the sss gene locus. qvr mutants were shown to have severely reduced Shaker (SH)-dependent K+ currents. The authors showed that qvr is in fact an allele of sss, and that SH protein levels are substantially reduced in sss P1 flies, indicating that SSS is an important regulator of the SH-dependent K+ channel.

Together these findings indicate that SSS is an important regulator of homeostatic sleep. The authors propose that SSS might link homeostatic sleep drive to neuronal excitability by regulating the amount and activity of SH-dependent K+ channels. Although no sss homologue has been identified in vertebrates, functional homologues might be found in the future.