The extracellular loops of Smoothened play a regulatory role in control of Hedgehog pathway activation

Abstract

The Hedgehog (Hh) signaling pathway plays an instructional role during development, and is frequently activated in cancer. Ligand-induced pathway activation requires signaling by the transmembrane protein Smoothened (Smo), a member of the G-protein-coupled receptor (GPCR) superfamily. The extracellular (EC) loops of canonical GPCRs harbor cysteine residues that engage in disulfide bonds, affecting active and inactive signaling states through regulating receptor conformation, dimerization and/or ligand binding. Although a functional importance for cysteines localized to the N-terminal extracellular cysteine-rich domain has been described, a functional role for a set of conserved cysteines in the EC loops of Smo has not yet been established. In this study, we mutated each of the conserved EC cysteines, and tested for effects on Hh signal transduction. Cysteine mutagenesis reveals that previously uncharacterized functional roles exist for Smo EC1 and EC2. We provide in vitro and in vivo evidence that EC1 cysteine mutation induces significant Hh-independent Smo signaling, triggering a level of pathway activation similar to that of a maximal Hh response in Drosophila and mammalian systems. Furthermore, we show that a single amino acid change in EC2 attenuates Hh-induced Smo signaling, whereas deletion of the central region of EC2 renders Smo fully active, suggesting that the conformation of EC2 is crucial for regulated Smo activity. Taken together, these findings are consistent with loop cysteines engaging in disulfide bonds that facilitate a Smo conformation that is silent in the absence of Hh, but can transition to a fully active state in response to ligand.

Fig. 1.

Conserved loop cysteines play a regulatory role in Smo signaling. (A) Alignment of Drosophila residues 306-596 against sequences of zebrafish, mouse, and human Smo proteins and Drosophila Frizzled2. Transmembrane residues are light gray, intracellular loops are dark gray and extracellular loops are black. Conserved loop 1 (L1) C320 and C339, loop 2 (L2) C413 and loop 3 (L3) C513 and C525 cysteines (Drosophila numbering) are indicated by arrowheads. Alignments were generated using StrapAlign (http://www.bioinformatics.org/strap/). (B) EC1 and EC2 cysteine mutants differentially rescue smo knockdown. Cl8 cells were treated with control or smo 5′UTR dsRNA and transfected with the indicated pAc-myc-smo expression vector. The ability of wild type or each of the Myc-Smo C to A mutants to rescue ptc-luciferase activity in presence of control vector (light gray bars) or pAc-hh (dark gray bars) is shown. The Hh-induced level of activity obtained in the presence of control dsRNA was set to 100%, and percent reporter activity relative to this value is shown. Activity levels are normalized to a pAc-renilla transfection control. Error bars indicate standard error of the mean (s.e.m.). (C) Loop 1 C to A mutants are dominant positive. Myc-Smo C to A mutant proteins were expressed in Cl8 cells with or without Hh, as indicated. Reporter activity was assessed as in B. In each case, 1X corresponds to 50 ng pAc-myc-smo. Error bars indicate s.e.m. (D) EC1 and EC2 cysteine mutants differentially activate downstream effectors. Lysates of Cl8 cells expressing wild-type or C to A mutant Smo proteins in the presence of pAc-hh (+) or empty vector (–) were examined by western blot. Short (Fu) and long (Fu dark) exposures are shown to visualize phosphorylation-induced mobility shifts (lanes 5 and 7). Samples are normalized to protein. Kinesin (Kin) serves as a loading control. (E) EC3 C to A mutants do not induce ligand-independent activation of downstream effectors. Lysates of Cl8 cells transfected with wild-type or C to A mutant pAc-myc-smo expression vectors with pAc-hh (+) or empty vector (–) were examined by western blot. Samples are normalized to protein. Kin serves as a loading control. (F) C to A mutants have free cysteines. Membrane fractions prepared from cells transfected with the indicated smo constructs were treated with maleimide-PEG₁₁-biotin to label free cysteines. Biotinylated proteins were precipitated on NeutrAvidin agarose and surveyed by western blot against Smo (bottom panel). Wild-type and mutant Smo proteins demonstrate similar mobility before treatment (upper panel, black arrow). C to A mutants in affinity complexes migrate more slowly than wild-type Smo (bottom panel, gray arrow compared with black).

Fig. 2.

Smo EC1 and EC2 mutants demonstrate altered subcellular localization. (A-D) Cl8 cells expressing the indicated Smo constructs in the presence of empty vector (Hh–) or pAc-hh (Hh+) were analyzed by indirect immunofluorescence. Smo was detected with anti-Myc (red). Calreticulin-GFP-KDEL (green) marks the ER.

Fig. 3.

EC1 and EC2 mutants alter Hh signaling in vivo. (A-E) myc-smo constructs were expressed in FLP-out clones, as indicated. Wing imaginal discs from late third instar larvae were immunostained for Ci (green), Ptc (white) or Myc-Smo (red). All images were collected at equal gains. All discs are oriented anterior to the left, posterior to the right, dorsal up and ventral down. Clones are indicated by arrows.

Fig. 4.

Smo EC mutants induce wing phenotypes. (A-E) myc-smo constructs were expressed under control of C765-Gal4 at 25°C, as indicated. Wings from adult male flies were mounted on glass slides and imaged under equal magnification. For each cross, multiple flies were analyzed. Representative wings are shown. Driver wing serves as control (A).

Fig. 5.

C to A compound mutants reveal a pivotal role for EC2. (A) Activating double C to A Smo mutants induce robust reporter gene activation. myc-smo C to A double mutant constructs were expressed in wild-type or smo knockdown backgrounds (UTR dsRNA), as indicated, and assessed for their ability to activate the ptc-luciferase reporter construct. Fifty ng of the indicated pAc-myc-smo vector was used in dsRNA rescue transfections, while 100 ng of the indicated pAc-myc-smo vectors were used to test for dominant activity in the wild-type smo background. Activity is shown as percent activity relative to the control Hh response, set to 100%. All values are normalized to a pAc-renilla transfection control. Error bars indicate s.e.m. (B) Activating double C to A Smo mutants signal to downstream effectors. Lysates from Cl8 cells transfected with the indicated Myc-Smo double mutant vectors were examined by western blot. Anti-Myc was used to blot for Myc-Smo protein. Samples are normalized to protein concentration. Kin serves as loading control. (C) Activating C to A double mutants have altered subcellular localization. The indicated loop double C to A mutant constructs were transfected into Cl8 cells in the absence of Hh and examined for subcellular localization by indirect immunofluorescence. Smo (anti-Myc) is in red, Calreticulin-GFP-KDEL is green. (D) ΔEC2 induces robust reporter gene activity. Cl8 cells were treated with control or smo 5′UTR dsRNA and transfected with the indicated pAc-myc-smo expression vector. Activity is shown as percent activity relative to the control Hh response. Error bars indicate s.e.m. (E) ΔEC2 signals to downstream effectors. Lysates from Cl8 cells transfected with the indicated pAc-myc-smo constructs were examined by western blot. Anti-Myc was used to blot for Myc-Smo. Samples are normalized to protein concentration. Kin serves as loading control.

Fig. 6.

The TM3-EC2 bond is important for mammalian Smo. (A) TM3-EC2 mutants are constitutively active. NIH3T3 cells were transfected with the indicated pCDNA-mSmo construct in the presence of pCDNA-Shh or empty vector. Relative activation of 8XGlibs-luc, normalized to pRL-TK is shown. Assays were repeated twice in triplicate and all data pooled. Error bars indicate s.e.m. (B-D) The TM3 C to A mutant is retained in the ER. NIH3T3 cells were transfected with the indicated pCDNA-mSmo expression vector in the absence (B-C) or presence (D) of pCDNA-Shh, and examined for Smo subcellular localization by indirect immunofluorescence. (B-C) Smo is green, GRP94 (ER marker) is red. (D) Smo is green, γ-tubulin (red) marks the base of the primary cilium and DAPI (blue) marks the nucleus. Whereas wild-type Smo enters the primary cilium in the presence of Shh, the C318A Smo mutant does not.