In keeping with its namesake, Hedgehog (Hh) — or, at least, components of the Hedgehog pathway — has come out of hibernation. Thanks to work by Phil Beachy and colleagues, reported in Science, Dally-like protein (Dlp) and casein kinase 1α (CK1α) have newly assigned roles in Hh signalling.

Because classical genetic approaches to studying Hh signalling have their limitations — not least because the pathway functions during embryogenesis — the authors chose to use RNA interference (RNAi) to disrupt gene function and they combined this approach with a cultured cell assay to screen Drosophila genes. The assay involved transfecting cells derived from the wing imaginal disc (cl-8 cells) with a Hh pathway-responsive luciferase reporter and double-stranded RNA (dsRNA) for RNAi. The assay used exogenous Hh ligand, so it rules out components that are involved in Hh synthesis or distribution, and is also quantitative. RNAi targeting of the positive regulatory elements of the pathway — Smoothened (Smo), Fused (Fu) and Cubitus interruptus (Ci) — inhibited the response to Hh, whereas targeting negative elements such as Costal 2 (Cos2) and Patched (Ptc) activated or enhanced the response. The assay also allowed the study of gene interactions and epistasis, as several dsRNA species could be combined.

To test the system, the authors used a library containing dsRNAs that corresponded to all of the kinases and phosphatases predicted from the Drosophila genome sequence. Several dsRNA pools that affected reporter activity were identified and, when re-screened, three dsRNAs emerged. Of these targets, CK1α had not previously been implicated in Hh signalling.

Extending the screen to a dsRNA library based on the Drosophila Gene Collection Release 1 — which contains cDNAs corresponding to ∼43% of the genes predicted from the Drosophila genome sequence — identified four gene targets that had not previously been implicated in Hh signalling. CK1α was one of these; another was Dlp, a glycipan heparin sulphate proteoglycan. Both are thought to function in the Wingless (Wg) signalling pathway.

Immunostaining showed that Dlp localized to the surface of cl-8 cells. RNAi of Ptc suppressed the requirement for Dlp in the Hh response, which indicates that Dlp could function upstream or at the level of Ptc. As Ptc is a membrane protein, Dlp might help deliver Hh to Ptc by concentrating Hh on the cell surface and, indeed, preliminary evidence showed that Dlp associates with Hh. Consistent with such a role, when Hh was expressed in a membrane-anchored form, RNAi of Dlp didn't block the signal response, which indicates that Dlp might function normally to concentrate the Hh signal.

Indications from the library screens were that CK1α might control basal pathway activity, as dsRNA increased basal reporter activity. Ci was required for this increase, whereas Smo and Fu were not, implying that CK1α is upstream or at the level of Ci, but downstream of Smo and Fu.

As a regulator of the basal activity of Hh and Wg signalling pathways, CK1α could function as a tumour suppressor in several cancers that are associated with overactivation of one or the other pathway. Similarly, deletion of a human chromosomal region that contains GPC6, a mammalian glycipan member that is closely related to Dlp, is associated with many human malformations. So, too, are mutations in one of the two remaining genes identified in the screen that had not previously been implicated in Hh signalling. Hopefully, discovering the functions of these genes in the Hh pathway won't be too prickly a problem!