This is not the first strange episode in Hh biology. Hh looks like a bacterial cell wall protein, and uses a bizarre mechanism related to that of self-splicing proteins to attach cholesterol to itself. “This signaling pathway, which has very profound roles in multicellular organisms, was very clearly put together with bits of this and that,” says Beachy. “At this point, nothing surprises us.”
Ptc was thought to be a conventional receptor that gripped Smo in an inhibitory embrace. But Taipale suspected that the original coimmunoprecipitation data were tainted by overexpression and the hydrophobicity of Ptc and Smo. “We thought we should start from scratch,” he says.
Taipale failed to find a significant interaction between the two proteins, and like others found that the proteins were present in different parts of the cell. Furthermore, free Ptc (in excess to Smo) affected pathway activity, and substoichiometric Ptc (1:45 of Ptc:Smo) resulted in 80% reduction of Smo activity, suggesting that Ptc acts catalytically.
Ptc is similar to bacterial proton-driven transporters. Beachy's group found that mutation of two channel-conserved residues led to a dramatic defect in Ptc activity. Thus, Hh may block Ptc from shipping in a Smo inhibitor or shipping out a Smo activator.
The relevant inhibitor or activator is unknown, but it may resemble cyclopamine. This steroidal alkaloid was first discovered when some Idaho sheep munched a maize lily and had malformed cyclopic offspring. The chemical culprit, cyclopamine, was isolated. Some thirty years later, Beachy found that mice with Hh defects suffered a similar fate, and that cyclopamine inhibits Smo.
Beachy has now found that cyclopamine regresses murine medulloblastomas, probably because these brain tumors result from Hh-dependent proliferation of stem cells. If other similarly aggressive but more common tumor types are found to be dependent on Hh, then cyclopamine may be the basis for an important cancer drug. ▪