Smoothened adopts multiple active and inactive conformations capable of trafficking to the primary cilium

Abstract

Activation of Hedgehog (Hh) signaling requires the transmembrane protein Smoothened (Smo), a member of the G-protein coupled receptor superfamily. In mammals, Smo translocates to the primary cilium upon binding of Hh ligands to their receptor, Patched (Ptch1), but it is unclear if ciliary trafficking of Smo is sufficient for pathway activation. Here, we demonstrate that cyclopamine and jervine, two structurally related inhibitors of Smo, force ciliary translocation of Smo. Treatment with SANT-1, an unrelated Smo antagonist, abrogates cyclopamine- and jervine-mediated Smo translocation. Further, activation of protein kinase A, either directly or through activation of Galphas, causes Smo to translocate to a proximal region of the primary cilium. We propose that Smo adopts multiple inactive and active conformations, which influence its localization and trafficking on the primary cilium.

Figure 1. Hh pathway agonists and antagonists stimulate Smo translocation to the primary cilium.

(A) Wild-type mouse embryonic fibroblasts (MEFs) stained with antibodies against endogenous Smo (green), acetylated (AC) tubulin (red), and DAPI (blue). MEFs treated with ShhN-conditioned media (ShhN-CM), purmorphamine, cyclopamine, and jervine display Smo staining along the entire length of the cilium. Note that fixation with paraformaldehyde produces artifactual nuclear staining with the Smo antibody; this nuclear staining is not visible after fixation in methanol (See Fig. 3C). (B) Quantification of the percentage of Smo-positive (Smo⁺) cilia in MEFs after treatment with ShhN CM and cyclopamine for the indicated times. Cyclopamine induces Smo translocation to a significant number of cilia after 6 hours of treatment, although this is less than that induced by ShhN CM. All quantification of cilia staining in this and subsequent figures represents at least two separate experiments with a minimum of 100 cilia scored per time point and condition. Error bars indicate +/− standard deviation (SD). (C) Quantification of Smo⁺ cilia in MEFs after treatment with indicated compounds for 24 hours. 20-α-hydroxysterol (20α-OHC), purmorphamine, cyclopamine, and jervine induce Smo translocation to the cilium. Error bars indicate +/− SD.

Figure 2. SANT-1 inhibits cyclopamine-and jervine- induced Smo translocation to the primary cilium.

(A) Quantification of Smo⁺ cilia in wild-type MEFs after treatment with indicated compounds for 24 hours. SANT-1 inhibits cyclopamine- and jervine-induced ciliary translocation of Smo. Error bars indicate +/− SD. (B) Fold activation of 8xGliBS-luciferase Hh reporter in wild-type MEFs after agonist and antagonist treatment for 48 hours. Treatment with ShhN-CM, 20-α-hydroxysterol (20α-OHC), or purmorphamine [but not 7β-hydroxysterol (7β-OHC), pro-vitamin D3, 7-dehydrocholesterol (7-DHC), cyclopamine, jervine or SANT-1] activated a Gli transcriptional response. Results are representative of three experiments in two wild-type MEF lines, and were normalized to a constitutively active Renilla luciferase reporter. Error bars indicate +/− SD. (C) Quantification of Smo⁺ cilia in Ptch1 ^−/− MEFs after treatment for 24 hours. Cyclopamine and jervine do not disrupt constitutive Smo localization. SANT-1 removes Smo from the cilium of Ptch1 ^−/− MEFs in the presence or absence of cyclopamine and jervine. Error bars indicate +/− SD. (D) Percentage activity of 8xGliBS-luciferase reporter in Ptch1 ^−/− MEFs after treatment for 48 hours. Cyclopamine, jervine or SANT-1 efficiently block Hh responses. Note that 7-DHC and pro-vitamin D3 do not significantly inhibit the activity of the reporter. Results are the mean of three independent experiments, and were normalized to a constitutively active Renilla luciferase reporter. Error bars indicate +/− SD. (E) Ptch1 ^−/− MEFs stained with antibodies against acetylated (AC) tubulin (red), Smo (green) and DAPI (blue). SANT-1 removes Smo from the cilium of Ptch1 ^−/− MEFs in the presence or absence of cyclopamine.

Figure 3. Modulation of Gαs and protein kinase A (PKA) causes Smo accumulation in a proximal region of the primary cilium.

(A) Treatment of wild-type MEFs with cholera toxin (CTX) for 24 hours induces Smo (green) translocation to the cilium (red). (B) Quantification of CTX- and FSK-induced Smo ciliary translocation after treatment for the indicated times. Treatment of wild-type MEFs with pertussis toxin (PTX) does not stimulate ciliary translocation of Smo. Error bars indicate +/− SD. (C) Wild-type MEFs were fixed in methanol and stained with antibodies against γ-tubulin (red; label the basal body) and Smo (green). In contrast to the relatively uniform Smo staining on the cilium induced by ShhN, CTX treatment causes Smo accumulation proximal to the basal body. (D) 8xGliBS-luciferase Hh reporter activity in wild-type MEFs with the indicated treatment. CTX and FSK do not activate the Hh pathway despite inducing ciliary localization of Smo. Data are representative of three independent experiments. Hh reporter activity was normalized to β-galactosidase activity produced from a constitutive hsp68-lacZ reporter, as alteration of PKA affected Renilla luciferase activity (data not shown). Error bars indicate +/− SD.

Figure 4. ShhN ligand and SANT-1 act dominantly to PKA-induced Smo translocation.

(A) Wild-type MEFs treated with indicated compounds for 24 hours stained with antibodies against acetylated (AC) tubulin (red), Smo (green) and DAPI (blue). ShhN causes Smo distribution along the length of the cilium when PKA is activated, whereas SANT-1 inhibits any type of Smo trafficking on the cilium. (B) Quantification of Smo⁺ cilia in wild-type MEFs after treatment for 24 hours. Nearly all cilia are Smo⁺ when treated with ShhN and CTX or FSK. Error bars indicate +/− SD. (C) Quantification of Smo⁺ cilia in wild-type MEFs after 24 hours. SANT-1 inhibits CTX- and FSK-mediated stimulation of Smo translocation to the proximal region of the primary cilium. Error bars indicate +/− SD. (D) 8xGliBS-luciferase reporter activity in wild-type MEFs with indicated treatments for 48 hours. CTX and FSK, but not PTX inhibit ShhN stimulation of the reporter. Normalization was performed as in Figure 3. Error bars indicate +/− SD.