Structural Basis of Smoothened Activation in Hedgehog Signaling

Abstract

Summary The seven transmembrane-spanning protein Smoothened is the central transducer in Hedgehog signaling, a pathway fundamental in development and cancer. Smoothened is activated by cholesterol binding to its extracellular cysteine-rich domain (CRD). How this interaction leads to changes in the transmembrane domain and Smoothened activation is unknown. Here, we report crystal structures of sterol-activated Smoothened. The CRD undergoes a dramatic reorientation, allosterically causing the transmembrane domain to adopt a conformation similar to active G protein-coupled receptors. We show that Smoothened contains a unique inhibitory π-cation lock, which is broken upon activation and is disrupted in constitutively active oncogenic mutants. Smoothened activation opens a hydrophobic tunnel, suggesting a pathway for cholesterol movement from the inner membrane leaflet to CRD. All Smoothened antagonists bind the transmembrane domain and block tunnel opening, but cyclopamine also binds the CRD, inducing the active transmembrane conformation. Together, these results define the mechanisms of Smoothened activation and inhibition.

A NanoBRET-based binding assay for Smoothened allows real time analysis of small-molecule ligand binding and distinction of two separate ligand binding sites for BODIPY-cyclopamine

“…Here, we have been able to pharmacologically separate two BODIPY-cyclopamine binding sites on the 7TM core of SMO. Total BODIPY-cyclopamine binding to full length Nluc-SMO is composed of at least two components: ligand binding to the CRD (Huang et al, 2016;Huang et al, 2018)for which we did not find evidence for in our experimentsand the 7TM core (Huang et al, 2018;Weierstall et al, 2014). Most importantly, SANT-1, which solely binds to the receptor core in the lower pocket of the SMO binding site , competes with BODIPY-cyclopamine more efficiently in the CRD Nluc-SMO compared to the full length receptor.…”

“…The development of a NanoBRET-based ligand binding assay provides an interesting complement to GPCR pharmacology enabling ligand binding studies on living cells in real time with simplified protocols (Stoddart et al, 2015;Stoddart, Kilpatrick & Hill, 2018 Recent insight into the molecular mechanisms of drug action on SMO by crystallography and CryoEM provide somewhat controversial yet intriguing information regarding cyclopamine and cholesterol interaction with the 7TM ligandbinding site and the CRD (Deshpande et al, 2019;Huang et al, 2016;Huang et al, 2018;Qi, Liu, Thompson, McDonald, Zhang & Li, 2019;Weierstall et al, 2014). Here, we have been able to pharmacologically separate two BODIPY-cyclopamine binding sites on the 7TM core of SMO.…”

“…These and other ligands, such as SMO agonists (e. g. SAG1.3, purmorphamine), neutral antagonists (e. g. cyclopamine, SANT-1) or inverse agonists (e. g. cyclopamine-KAAD), are frequently used to explore SMO pharmacology (Chen, 2016;Chen, Taipale, Young, Maiti & Beachy, 2002;Rominger et al, 2009). To date, binding affinities of these compounds were determined using classical radioligand binding methods (Frank-Kamenetsky et al, 2002;Rominger et al, 2009;Wang et al, 2014) and, more often, fluorescence-based assays using for example the greenyellow fluorescent BODIPY-cyclopamine (Chen, Taipale, Cooper & Beachy, 2002;Chen, Taipale, Young, Maiti & Beachy, 2002;Gorojankina et al, 2013;Huang et al, 2016;Huang et al, 2018;Manetti et al, 2010). The fluorescently-labelled cyclopamine has been used in three separate experimental approaches: assessment of ligand-receptor interaction based on detection of fluorescence using confocal microscopy, flow cytometry or fluorescence polarization (Bee et al, 2012;Chen, Taipale, Cooper & Beachy, 2002;Chen, Taipale, Young, Maiti & Beachy, 2002;Gorojankina et al, 2013;Huang et al, 2016;Huang et al, 2018).…”

“…To date, binding affinities of these compounds were determined using classical radioligand binding methods (Frank-Kamenetsky et al, 2002;Rominger et al, 2009;Wang et al, 2014) and, more often, fluorescence-based assays using for example the greenyellow fluorescent BODIPY-cyclopamine (Chen, Taipale, Cooper & Beachy, 2002;Chen, Taipale, Young, Maiti & Beachy, 2002;Gorojankina et al, 2013;Huang et al, 2016;Huang et al, 2018;Manetti et al, 2010). The fluorescently-labelled cyclopamine has been used in three separate experimental approaches: assessment of ligand-receptor interaction based on detection of fluorescence using confocal microscopy, flow cytometry or fluorescence polarization (Bee et al, 2012;Chen, Taipale, Cooper & Beachy, 2002;Chen, Taipale, Young, Maiti & Beachy, 2002;Gorojankina et al, 2013;Huang et al, 2016;Huang et al, 2018). While these methods offer valuable insight into ligand-receptor binding in live cells, they suffer from several shortcomings, including: 1) laborious protocols that require long ligand incubation times (up to 4-5 hours), 2) extensive cell washing due to lipophilicity of BODIPY-cyclopamine, 3) high levels of non-specific binding in untransfected cells or in the presence of saturating concentrations of unlabeled competitors and 4) the need for data normalization of BODIPY-fluorescence values to receptor expression values.…”

Non-acylated Wnts Can Promote Signaling

“…It is possible that the acyl chains that bind FZD CRDs are similarly distinct from those attached to Wnt molecules – themselves bound primarily through site 2 interactions (Figure 1). Free fatty acid binding to FZDs may reflecting an as-yet-unappreciated regulatory role for lipid-like molecules in Wnt signaling, paralleling recent discoveries with Smoothened (Huang et al, 2018). …”