) gene was identified two decades ago in Drosophila
as a critical regulator of cell-fate determination during embryogenesis [1
]. Subsequent work in several model systems has defined and characterized the Hh
gene family that encodes highly conserved secreted signaling proteins (for review see [2
]). Hedgehog (Hh) proteins are synthesized as approximately 45 kDa precursors that autoprocess in an unprecedented fashion, resulting in the covalent attachment of a cholesterol moiety to the amino-terminal half of the precursor [2
]. This processed amino-terminal domain, Hh-Np, is responsible for the activation of a unique and complex signaling cascade that is essential for controlling cell fate throughout development and into adulthood [2
]. In mammals there are three Hh-family proteins: Sonic (Shh), Indian (Ihh), and Desert (Dhh). Gene-targeting experiments in mice have demonstrated that the development and patterning of essentially every major organ requires input from the Hh pathway [2
culture systems of neuronal tissues have been used to characterize the biology of the Hh-signaling pathway. Most notably, the neural-plate explant assay has defined the concentration-dependent role that ventrally expressed Shh plays in opposing dorsally expressed bone morphogenetic proteins (BMPs) to pattern the neural tube [2
]. The assay demonstrates that the Hh-signaling cascade can distinguish between small concentration differences in the Hh ligand to instruct the differentiation of specific neuronal cell types. Additional insights have been gained by utilizing cultures of postnatal cerebellar neuron precursors [2
]. These studies have shown that Hh patterns the cerebellum by promoting proliferation of the granule neuron precursors. Given the role that Hh signaling plays in promoting progenitor-cell proliferation, it is not surprising that misregulation of Hh signaling has been implicated in the biology of certain cancers, in particular basal cell carcinoma (BCC) and medulloblastoma.
The Hh-signaling pathway comprises three main components: the Hh ligand; a transmembrane receptor circuit composed of the negative regulator Patched (Ptc) plus an activator, Smoothened (Smo); and finally a cytoplasmic complex that regulates the Cubitus interruptus (Ci) or Gli family of transcriptional effectors. Additional pathway components are thought to modulate the activity or subcellular distribution of these molecules [2
]. There is positive and negative feedback at the transcriptional level as the Gli1
genes are direct transcriptional targets of activation of the pathway.
Smo is a seven-pass transmembrane protein with homology to G-protein-coupled receptors (GPCRs), while Ptc is a twelve-pass transmembrane protein that resembles a channel or transporter. Consistent with its role as an essential pathway inhibitor, removal of Ptc renders the Hh pathway constitutively 'on', independent of the Hh ligand. Similarly, specific point mutations in the transmembrane helices of Smo are capable of constitutively stimulating the pathway, effectively bypassing Ptc inhibition [3
]. At present, a controversy surrounds the mechanism by which Ptc inhibits Smo. Although early studies suggested a simple, direct, stoichiometric regulation, more recent data support a more complicated indirect or catalytic model [2
]. And although it has been demonstrated that Hh directly interacts with [4
] and destabilizes [5
] Ptc, the downstream molecular events remain obscure. In particular, little is known about the means by which Ptc exerts its inhibitory effect on Smo, or how Smo communicates with the cytoplasmic Ci/Gli transcription factor complex.
Through a 'chemical genetic' approach of identifying and studying the mechanism of action of small-molecule agonists (and antagonists), we hoped to uncover some of the complexities of the Hh-signaling system. Small-molecule modulators of growth-factor pathways have proven valuable in providing enhanced understanding of the intracellular events that occur subsequent to receptor activation, and in establishing the biological functions of these pathways [6
]. In Hh signaling, multiple insights have been gained through the use of the plant-derived Hh antagonist cyclopamine [9
] and a recently identified synthetic small-molecule Hh-signaling inhibitor, Cur61414 [17
]. Interestingly, these specific inhibitors of Hh signaling appear to function downstream of Ptc but their precise molecular target(s) and mechanism of action are unknown.
Although genetic manipulations involving gain-of-function point mutations of Smo [3
] have demonstrated that the pathway can be activated independently of Hh ligand, no small molecules with this capability have been identified. Indeed, it has proven difficult to identify small-molecule agonists of any signaling pathway activated by a protein ligand. Two examples have recently been described, however. One involved identification of a non-peptide activator of the granulocyte colony-stimulating factor (GCSF) pathway that appeared to act via receptor oligomerization [18
]. Another report described a small-molecule activator of the insulin-signaling pathway that also acts at the level of the receptor [19
Since the Hh receptor, Ptc, serves to inhibit signaling, a small-molecule pathway activator would need to be capable of one of the following: first, interfering with the inhibitory effect that Ptc exerts on Smo; second, activating Smo without affecting Ptc; or third, activating the pathway downstream of Smo. Identifying small molecules with any of these activities would provide useful information concerning the details of Hh signaling and would also provide a simple means of modulating activity of the pathway in vivo or in vitro.
In this article, we show that a non-peptidyl small-molecule agonist of Hh signaling has been identified that has all the known signaling properties of the recombinant Hh protein. But this agonist, unlike Hh protein, appears to bypass the Ptc-regulatory step, by interacting directly with Smo. Furthermore, studies with the agonist and several antagonists of Hh signaling suggest that Smo can be activated or inhibited by direct interaction with small-molecule ligands. These observations suggest that the Ptc-Smo receptor circuit may incorporate native small-molecule ligands in the regulation of Hh signaling.