Hedgehog-regulated atypical PKC promotes phosphorylation and activation of Smoothened and Cubitus interruptus in Drosophila Kai Jianga, Yajuan Liua, Junkai Fana, Garretson Epperlya, Tianyan Gaoa,b, Jin Jiangc, and Jianhang Jiaa,b,1 a

Markey Cancer Center and bDepartment of Molecular and Cellular Biochemistry, University of Kentucky College of Medicine, Lexington, KY 40536-0509; and cDepartment of Developmental Biology, University of Texas Southwestern Medical Center, Dallas, TX 75390

Edited by Norbert Perrimon, Howard Hughes Medical Institute, Harvard Medical School, Boston, MA, and approved October 3, 2014 (received for review September 5, 2014)

Smoothened (Smo) is essential for transduction of the Hedgehog (Hh) signal in both insects and vertebrates. Cell surface/cilium accumulation of Smo is thought to play an important role in Hh signaling, but how the localization of Smo is controlled remains poorly understood. In this study, we demonstrate that atypical PKC (aPKC) regulates Smo phosphorylation and basolateral accumulation in Drosophila wings. Inactivation of aPKC by either RNAi or a mutation inhibits Smo basolateral accumulation and attenuates Hh target gene expression. In contrast, expression of constitutively active aPKC elevates basolateral accumulation of Smo and promotes Hh signaling. The aPKC-mediated phosphorylation of Smo at Ser680 promotes Ser683 phosphorylation by casein kinase 1 (CK1), and these phosphorylation events elevate Smo activity in vivo. Moreover, aPKC has an additional positive role in Hh signaling by regulating the activity of Cubitus interruptus (Ci) through phosphorylation of the Zn finger DNA-binding domain. Finally, the expression of aPKC is up-regulated by Hh signaling in a Ci-dependent manner. Our findings indicate a direct involvement of aPKC in Hh signaling beyond its role in cell polarity. aPKC

| Ci | Hh | Par6 | Smo

H

edgehog (Hh) was originally discovered as a segment polarity gene involved in Drosophila embryo development (1). It has been shown that Hh family members function as morphogens and play critical roles in pattern formation and cell growth control; therefore, aberrant Hh signaling causes birth defects as well as several types of cancer (2–4). The Hh signal is transduced by a signaling cascade that is highly conserved among different species. One of the best model systems for studying Hh signal transduction is the Drosophila wing imaginal disc, which is divided into posterior (P) and anterior (A) compartments. The P compartment cells express and secrete Hh proteins that act upon neighboring A compartment cells located adjacent to the A/P boundary to induce the expression of decapentaplegic (dpp) (5, 6). The Dpp protein then diffuses bidirectionally into both the A and P compartments and functions as a morphogen to control the growth and patterning of cells in the entire wing in a concentrationdependent manner (7–9). Hh also activates other genes, including collier (col), patched (ptc), and engrailed (en). Low levels of Hh are able to induce the expression of dpp, whereas higher levels of Hh are also able to activate col and ptc. The induction of en appears to require the highest doses of Hh signaling activities (10, 11). Cell polarity is established by a conserved protein complex containing the partition defective (Par) proteins, Bazooka (Baz)/ Par3 and Par6, as well as the atypical PKC (aPKC), all of which have been shown to be essential for the formation of polarized epithelia (12). Epithelial polarity also depends on polarity complexes containing Crumbs (Crb), which localizes to the apical domain and interacts with aPKC-Par6 in both Drosophila and mammalian epithelia, thereby acting as the apical determinant (13–15). In Drosophila embryo, the aPKC–Baz–Par6 apical protein complex functions together with Crb to regulate apical/basal E4842–E4850 | PNAS | Published online October 27, 2014

polarity, and this apical complex is antagonized by a basolateral protein complex containing Scribble (Scrib), Discs large (Dlg), and Lethal giant larvae (Lgl) (16, 17). It is unknown whether there is a genetic or physiological interaction between Hh signaling components and the epithelial polarity complex. Smoothened (Smo), an atypical G protein-coupled receptor, is essential in both insects and mammals for transduction of the Hh signal. Abnormal Smo activation results in basal cell carcinoma (BCC) and medulloblastoma, so it remains an attractive therapeutic target. In Drosophila, binding of Hh to the receptor complex consisting Ptc and Interference Hedgehog (Ihog) relieves the inhibition of Smo by Ptc (10, 18, 19), ultimately allowing Smo to activate the Cubitus interruptus (Ci)/Gli family of Zinc finger transcription factors, and thereby induce the expression of Hh target genes, such as dpp, col, ptc, and en (10, 11). Studies have shown that Hh induces cell surface accumulation and phosphorylation of Smo by multiple kinases, including PKA and casein kinase 1 (CK1), CK2, and G protein-coupled receptor kinase 2 (Gprk2), which activate Smo by inducing differential phosphorylation, and thus the conformational change in the protein (20–26). In addition, Smo cell surface accumulation is controlled by endocytic trafficking that is mediated by ubiquitination (27, 28). In mammals, Hh signal transduction depends on the primary cilium, and ciliary accumulation is required for Smo activation (29–32). Moreover, phosphorylation by multiple kinases promotes the ciliary localization of mammalian Smo Significance Hedgehog (Hh) signaling by Smoothened (Smo) is mediated by phosphorylation and cell surface/cilium accumulation, but how the localization of Smo is controlled remains poorly understood. We show that the atypical PKC (aPKC)–partition defective 6 (Par6) complex promotes Hh signaling by phosphorylating Smo and regulating Smo basolateral accumulation in addition to phosphorylating the transcription factor cubitus interruptus (Ci). Our results demonstrate direct involvement of aPKC in Hh signaling beyond its role in cell polarity and suggest that basolateral accumulation of Smo is critical for its activity. Abnormal activation of Smo results in several types of cancers, and Smo can easily acquire drug resistance through mutations. A better understanding of the mechanisms of Smo regulation is critical to developing more effective therapeutic treatments for cancers caused by Smo dysregulation. Author contributions: J. Jia designed research; K.J., Y.L., and J.F. performed research; T.G. and J. Jiang contributed new reagents/analytic tools; K.J., Y.L., J.F., G.E., T.G., J. Jiang, and J. Jia analyzed data; and J. Jia wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1

To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1417147111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1417147111

PNAS PLUS

(33). Therefore, cilium may function as a signaling center for the Hh pathway in mammals (34). However, the precise mechanism by which Smo cell surface or ciliary accumulation is regulated remains unclear. Here, we report the identification of aPKC as a positive regulator in the Hh pathway by phosphorylating Smo and regulating Smo accumulation in vivo. We found that the loss of function of aPKC blocks basolateral accumulation of Smo and Hh signal transduction. We then examined the activity of Smo variants with phosphomimetic or phosphodeficient mutations of aPKC and adjacent CK1 phosphorylated residues, and found that aPKCmediated phosphorylation of Smo positively regulates Smo activity. We further demonstrated that aPKC has an additional role in Hh signaling downstream of Smo by regulating the transcriptional activity of Ci through phosphorylation of the Zn finger DNAbinding domain. We found that Hh signaling up-regulates the protein level of aPKC in wing disc, indicating a positive feedback regulation of Hh components by aPKC. Finally, we showed that inactivation of other polarity proteins, with the exception of Par6, did not have any effect on Smo accumulation and Hh signal transduction, suggesting a direct role of aPKC in Hh signaling beyond its functions in cell polarity. Results

Jiang et al.

Fig. 1. Positive role of aPKC in Hh signaling. (A) WT wing disc immunostained to show Smo and col-lacZ expression. (B) Wing disc expressing UASaPKC-RNAi35001 with MS1096 Gal4 (a wing-specific Gal4 that has slightly higher Gal4 expression in the dorsal compartment than in the ventral compartment) was stained for Smo and Ci. (C–F) Wing discs expressing UASaPKC-RNAi35001 with MS1096 Gal4 were stained for col-lacZ, dpp-lacZ, ptclacZ, or PKC. (G) Wing disc expressing UAS-aPKCCAAX by MS1096 Gal4 was stained for Smo, dpp-lacZ, and Ci. Arrows indicate the elevation of Smo, ectopic dpp-lacZ, and Ci. (H) Wing disc expressing UAS-aPKCΔN by MS1096 Gal4 was stained for Smo and Ci. All wing imaginal discs shown in this study were oriented with anterior on the left and ventral on the top.

aPKC Acts Downstream of Ptc in Hh Pathway. We then determined the genetic epistasis between aPKC and Hh signaling components, including Ptc and Smo. We found that inactivation of Ptc by RNAi caused anterior expansion of col-lacZ expression, which was blocked by aPKC RNAi (Fig. 3 A–C), suggesting that aPKC acts downstream of Ptc. Interestingly, the levels of aPKC were elevated in ptc mutant cells of wing disc (Fig. 3D), suggesting that aPKC could be regulated by Hh signaling (see below). The elevated Smo staining in ptc mutant clones indicated the relief of Ptc inhibition on Smo (Fig. 3D). We also overexpressed aPKCΔN with MS1096 Gal4 in wing disc that contained smo mutant clones. Interestingly, the elevated Ci was diminished (Fig. 3E) and the expression of ptcLacZ at the A/P boundary was blocked (Fig. 3F) in smo mutant cells expressing aPKCΔN. These data suggest that Hh induces the expression of aPKC and that aPKC acts downstream of Ptc to regulate Smo in Hh signaling. aPKC Phosphorylates Smo and Promotes Smo Signaling Activity. Smo is phosphorylated at Ser680 and Ser683 in S2 cells in the presence of Hh-conditioned medium (20). However, the kinases responsible for Ser680 and Ser683 phosphorylation are unknown (Fig. 4A). Ser680 is a predicted PKC site, and Ser683 may be phosphorylated by either CK1 or GSK3, depending on the priming phosphoresidue. To determine the kinases involved in phosphorylating Ser680 and Ser683, we performed in vitro kinase assays using Smo656–755 purified by thrombin cleavage from GST-Smo fusion proteins. We found that Smo656–755 was phosphorylated by recombinant PKCζ, and the phosphorylation was enhanced by the addition of CK1 kinase (Fig. 4B, Upper). PKC inhibitor completely blocked, whereas the CK1 inhibitor CK1–7 partially blocked, the phosphorylation by two kinases PNAS | Published online October 27, 2014 | E4843

DEVELOPMENTAL BIOLOGY

ces Smo phosphorylation by PKA, CK1, CK2, and Gprk2. However, a previous mass spectroscopy study indicated that Smo is phosphorylated at 26 Ser/Thr residues in vivo (20), many of which are not linked to the aforementioned kinases, leading to the hypothesis that additional kinase(s) may be involved in Smo phosphorylation and regulation. We screened a partial RNAi library targeting the S/T kinases in the Drosophila genome. Individual RNAi lines from the Vienna Drosophila RNAi Center (VDRC) and Bloomington Stock Center (BSC) were expressed through wing-specific MS1096 Gal4 to determine whether they induced adult wing phenotypes or caused changes in Smo and Ci accumulation in wing discs (the results of the whole screen will be published elsewhere). We found that inactivation of aPKC by aPKC-RNAi35001 attenuated Hh-induced accumulation of Smo (Fig. 1B) and inhibited the expression of Hh target genes, such as col-lacZ, dpp-lacZ, and ptc-lacZ (Fig. 1 C–E). The phenotype of aPKC-RNAi is unlikely due to an off-target effect because expression of two different transgenic RNAi lines (VDRC v105624 and BSC 35001) targeting nonoverlapping regions of an aPKC sequence produced a similar phenotype and expression of either of these lines blocks aPKC expression in wing disc (Fig. 1F). To examine aPKC gain of function, we overexpressed a membranetethered form of aPKC (aPKCCAAX) or a constitutively active form of aPKC (aPKCΔN) lacking the N-terminal pseudosubstrate regulatory domain and found that gain of function of aPKC promoted Smo accumulation and induced ectopic dpp-lacZ expression and Ci accumulation (Fig. 1 G and H). We also used actin>CD2>Gal4 to generate clones that express aPKCCAAX and found aPKCCAAX expression resulted in Smo accumulation (Fig. S1A). To examine the physiological function of aPKC in Hh signaling further, we used mitotic recombination to generate clones homozygous for aPKCK06403, a strong allele of aPKC (35). We found that in aPKC mutant cells, Smo accumulated in apical membrane of wing disc (Fig. 2 A and C), leading to a reduction in basolateral Smo accumulation (Fig. 2 B and C). The expression of dpp-lacZ, en, and ptc was blocked in aPKC mutant cells (Fig. 2 B, D, and E). Consistently, inactivation of aPKC by RNAi blocked, whereas the expression of aPKCCAAX increased, basolateral accumulation of Smo (Fig. 2 F–H). These data suggest that aPKC regulates Hh signal transduction, at least in part, by promoting Smo basolateral accumulation.

CHEMISTRY

aPKC Positively Regulates Smo Basolateral Accumulation and Hh Signal Transduction. Previous studies have shown that Hh indu-

wing disc (Fig. S1B), indicating that Sgg/GSK3 does not regulate Smo accumulation in wing disc, and therefore that Ser683 is unlikely to be a GSK3 site. To examine whether aPKC phosphorylates Smo in cultured cells, we transfected Myc-SmoWT into S2 cells and treated cells with Hh-conditioned medium, together with aPKC dsRNA or GFP dsRNA. Hh induced slow-migration hyperphosphorylated forms of Myc-Smo, indicative of Smo phosphorylation (Fig. 4C, Upper, lane 2 compared with lane 1). We found that the Hhinduced Smo mobility shift was attenuated by aPKC RNAi but not by GFP RNAi (Fig. 4C, Upper, lanes 4 and 5), indicating that Smo phosphorylation was regulated by aPKC. RNAi of aPKC did not cause any change in Smo in the absence of Hh (Fig. 4C, Upper, lane 3 compared with lane 1). We further verified phosphorylation changes in Smo using the Phos-tag gel that specifically retards phosphorylated proteins (33, 38). We found that inactivation of aPKC by RNAi attenuated the phosphorylation of Smo, as indicated by the reduced mobility shift (Fig. 4C, Lower, and the corresponding densitometry graph on the right). To determine whether Hh induces Smo phosphorylation at Ser680 and Ser683, we transfected Myc-SmoWT or Myc-Smo680-3SA into S2 cells, followed by treatment with Hh-conditioned medium or control medium. As shown in Fig. 4D (Upper), the Hh-induced electrophoretic mobility shift of Myc-Smo was impaired by 6803SA mutation (Fig. 4D, Upper, lane 4 compared with lane 3). In addition, when S2 cells were treated with okadaic acid (OA, a phosphatase inhibitor), Myc-SmoWT existed in peak levels of phosphorylation, whereas Myc-Smo680-3SA showed a reduction in mobility shift (Fig. 4D, Upper, lane 6 compared with lane 5), in-

Fig. 2. Loss of aPKC prevents Smo accumulation on the cell surface and blocks Hh signaling. (A) Wing disc bearing aPKCK06403 mutant clones and expressing UAS-P35 (a cell death inhibitor that promotes the survival of mutant cells) by MS1096 Gal4 was stained for Smo and Ci. The image shows the apical layer. (B) Wing disc with the same genotype as in A was stained for Smo and dpp-lacZ. Arrows indicate the down-regulated Smo accumulation (gray), clones marked by the lack of GFP (green), and the blockade of dpp-lacZ expression in aPKC mutant cells near the A/P boundary. (C) Z-section was taken to determine Smo accumulation in aPKCK06403 mutant clone. The arrow indicates the increased Smo accumulation in the apical localization, and the arrowhead indicates the decreased Smo accumulation in the basolateral localization. Dashed lines mark the mutant clone that is defined by the lack of GFP expression. (D and E) Wing discs bearing the same genotype as in A were stained for En, Ptc, and Ci. Arrows indicate the blocked En and Ptc expression (gray) in cells lacking GFP expression (green). (F–H) WT wing disc or wing discs expressing either aPKCRNAi105624 or aPKCCAAX were stained for Smo. Arrows indicate the Smo basolateral localization. Dashed lines indicate the A/P boundary that is defined by Ci expression. All of the Z-scan sections were taken at 0.5 μm by confocal imaging (Olympus Fluoview, version1.7c) to monitor Smo localization in wing discs. All Z-sections shown in this study were oriented with the apical localization to the top.

(Fig. 4B, Upper, lanes 5 and 6). To examine further whether Ser683 phosphorylation is primed by phosphorylation at Ser680, we used Smo680SA and Smo683SA, which have a mutation at Ser680 and Ser683, respectively. The results showed that Ser683 was phosphorylated by CK1 when Ser680 was phosphorylated by recombinant PKCζ (Fig. 4B, Lower, lane 4), suggesting that Ser680 and Ser683 are likely PKC and CK1 sites, respectively. In addition, loss of shaggy (sgg) gene function resulted in accumulation of high levels of Ci (36, 37) but did not change Hh-induced Smo accumulation in E4844 | www.pnas.org/cgi/doi/10.1073/pnas.1417147111

Fig. 3. aPKC acts downstream of Ptc, and aPKC function requires Smo. (A– C) Wing discs expressing UAS-Ptc-RNAi28795 or UAS-aPKC-RNAi105624 alone or together by MS1096 Gal4 were stained for col-lacZ. Arrows point to the expression of col-lacZ. (D) Wing disc carrying ptc[s2] mutant clone was immunostained for PKC and Smo. Clones are marked by the lack of GFP expression. Arrows indicate the elevation of aPKC (red) and Smo (gray) by ptc mutation. (E and F) Wing discs carrying smo mutant clones and expressing UAS-aPKCΔN by MS1096 Gal4 were stained for Ci, PKC, and ptclacZ. Clones are marked by the lack of GFP expression. The arrow in E indicates the blocked Ci accumulation by mutating smo in the aPKC overexpression background. The arrow in F indicates the blockade of ptc-lacZ expression.

Jiang et al.

Jiang et al.

To test the functional effect of phosphorylation in vivo, we constructed Smo phosphomimetic or phosphodeficient mutants at Ser680 and Ser683. Smo transgenes were generated at a PNAS | Published online October 27, 2014 | E4845

DEVELOPMENTAL BIOLOGY

PNAS PLUS

dicating that mutating Ser680 and Ser683 attenuated Smo phosphorylation. This observation was confirmed by changes in mobility shifts of Myc-Smo on the Phos-tag gel (Fig. 4D, Lower).

CHEMISTRY

Fig. 4. Smo phosphorylation and activation by aPKC. (A) Schematic drawing of full-length Smo with the sequences of the three phosphorylation clusters shown underneath. The aPKC site and adjacent CK1 site are marked by an asterisk. Red, PKC site; blue, PKA site; green, CK1 site. (B) Autoradiograph of the in vitro kinase assays. (Upper) Smo proteins were incubated with recombinant PKCζ, recombinant CK1, or both in the presence of [γ-32P]ATP (lanes 2–4) or treated with PKC inhibitor or CK1 inhibitor (lanes 5 and 6). (Lower) Smo proteins were incubated with recombinant PKCζ in the presence of [γ-32P]ATP (lanes 1–3) or treated with recombinant CK1 in the presence of [γ-32P]ATP (lanes 4–6) following incubation with PKC in the presence of cold ATP (indicated by an asterisk). Equal amounts of Smo, Smo680SA, and Smo683SA purified from GST-Smo fusion proteins were used in these experiments. (C) S2 cells were transfected with Myc-SmoWT, followed by treatment with Hh-conditioned medium and the indicated dsRNA targeting aPKC or GFP. Cell extracts were immunoprecipitated with the anti-Myc antibody and Western blotted with the anti-Myc antibody to examine the mobility shift of Smo. The efficiency of GFP RNAi (Left Middle) and aPKC RNAi (Right Upper), respectively, was determined by Western blot (WB) analysis. GFP served as a transfection and loading control. β-Tubulin served as a loading control. (Left Lower) Phos-tag gel was also used to monitor the mobility shift of Sm, in which the immunoprecipitation (IP) products were achieved from a different batch of cells, whereby GFP RNAi and aPKC RNAi were also monitored. (Right Lower) Densitometry graph of Myc-Smo based on the Phos-tag gel is shown. (D) S2 cells were transfected with Myc-SmoWT or Myc-Smo680-3SA and treated with HhN-conditioned medium or 50 nM OA. Cell extracts were immunoprecipitated with the anti-Myc antibody, followed by Western blotting with the anti-Myc antibody to examine the mobility shift of the protein on regular gel (Upper) or on a Phos-tag gel (Lower). GFP served as a transfection and loading control. (E) WT wing was immunostained to show dpp-lacZ expression at the A/P boundary. (F–O) Wing discs expressing the indicated constructs at the VK5-attP locus by MS1096 Gal4 were stained for dpp-lacZ. Arrows in F and G indicate the ectopic dpp-lacZ expression induced by the overexpression of Smo. (P–R) Wing discs expressing the indicated constructs by MS1096 Gal4 were stained for En. (S) Activity of each individual construct was assayed by the ptc-luc reporter using S2 cells cotransfected with tub-Ci.

specific gemonic locus to ensure the same level of expression without positional effects (26). Expression of SmoWT induced a low level of ectopic dpp-lacZ expression (26) (Fig. 4F compared with WT dpp-lacZ expression in Fig. 4E). Single-site S>A mutation (Smo680SA; Fig. 4A) had less ectopic activity (Fig. 4G), whereas single-site S>D mutation (Smo680SD) had slightly higher activity compared with SmoWT (Fig. 4H). Expression of Smo680-3SD induced much higher levels of ectopic dpp-lacZ expression (Fig. 4J), whereas expression of Smo680-3SA had little, if any, activity (Fig. 4I). These data suggest that phosphorylation at Ser680 and Ser683 promotes Smo activation. Hh gradients promote Smo differential phosphorylation in a dose-dependent manner, and among the three PKA and CK1 phosphorylation clusters, phosphorylation of the membrane proximal cluster is more critical than phosphorylation of the distal cluster (26). In this study, we found that mutating the PKA site of the second cluster in Smo680-3SD (Smo 680-3SD-PKA2; Fig. 4A) blocked the ectopic activity (Fig. 4K compared with Smo680-3SD in Fig. 4J). In addition, SmoDAD, in which PKA and CK1 sites in the second cluster are mutated to Ala (Fig. 4A), barely had any ectopic activity even though the first and third clusters have S>D mutations (26) (Fig. 4L). SmoDAD680-3SD did not have any ectopic activity either (Fig. 4M). In contrast, SmoDSD680-3SD, keeping the second cluster as WT, induced potent dpp-lacZ expression (Fig. 4O compared with Fig. 4N). Phosphomimetic mutation at the three clusters (SmoDDD) led to hyperactivation of Smo, which induced peak levels of Hh signaling activity indicated by potent en expression (22, 26) (Fig. 4P). Furthermore, SmoDDD680-3SA and SmoDDD680-3SD, in which Ser680 and Ser683 were mutated either to Ala or Asp in the background of SmoDDD (Fig. 4A), induced en expression at comparable levels to SmoDDD (Fig. 4 Q–R). These data suggest that, phosphorylation at the second cluster of PKA/ CK1 sites is required for aPKC-mediated phosphorylation to activate Smo. We also performed a ptc-luciferase (luc) reporter assay in S2 cells, which were more sensitive than the wing discs, to examine the activity of Smo mutants (26). In this assay, we found that Smo680SD has higher activity than SmoWT (Fig. 4S). In addition, SmoDDD, SmoDDD680-3SA, and SmoDDD680-3SD induced comparable ptc-luc activity in cultured cells (Fig. 4S), which is consistent with the induced comparable levels of en expression in wing discs (Fig. 4 P–R). Also consistent with the data from imaginal discs, Smo680-3SD-PKA2 had much lower luciferase activity compared with Smo680-3SD, and SmoDAD680-3SD had much lower luciferase activity compared with SmoDSD680-3SD (Fig. 4S), confirming that phosphorylation at the second cluster of PKA/CK1 sites is required for aPKC to activate Smo. Additional Role of aPKC in Hh Signaling. Overexpression of aPKCΔN

caused Ci elevation (Fig. 1H), which was blocked by mutating smo (Fig. 3E), suggesting aPKC requires Smo to promote Hh signaling. Surprisingly, knockdown of aPKC by RNAi attenuated the ectopic dpp-lacZ expression induced by Smo680-3SD (Fig. 5 A and B). In addition, aPKC RNAi attenuated the activity of the superactive form of Smo, SmoDDD680-3SD (Fig. 5E compared with Fig. 5D). One possibility is that aPKC may phosphorylate Smo at additional residues. However, phosphomimetic or phosphodeficient mutation of other predicted PKC sites in the C terminus of Smo (including Ser885, Ser888, Ser890, Ser893, Ser957, and Thr1020) did not change the activity of Smo in wing discs. Another possibility is that aPKC regulates Smo through an additional role beyond phosphorylation, or through regulating Hh signaling component(s) downstream of Smo. In support of this hypothesis, we found that aPKC RNAi suppressed the activity of Ci−3P, in which three PKA sites in the phosphorylation clusters were mutated resulting in an active form of Ci (39) (Fig. 5 F and G). To determine further how aPKC regulates the activity of Ci, we examined whether aPKC regulates Ci through phosphoryE4846 | www.pnas.org/cgi/doi/10.1073/pnas.1417147111

lating its DNA-binding domain, as is the case for the regulation of Gli1 by aPKC (40). The expression of Ci-GA2, the Zn finger Ci DNA-binding domain (CiDB) fused to the Gal4 activation domain (41), induced ptc-lacZ expression mostly in P compartment cells (Fig. 5H). We found that Ci-GA2–induced ptc-lacZ expression was inhibited by RNAi of aPKC (Fig. 5I). In contrast, Ci-GA2 activity was enhanced by coexpression of aPKCCAAX, as indicated by the ectopic ptc-lacZ expression in A compartment cells (Fig. 5J). Sequencing analysis identified Thr512 and Thr590 in the Zn finger domain, which match the phosphorylation consensus sites for PKA and aPKC (Fig. 5O). However, coexpression of MC*, the constitutively active form of PKA, did not change the activity of Ci-GA2 (Fig. 5K), suggesting that these residues are likely to be aPKC sites. Thr512 and Thr590 were then mutated either alone or together in the background of Ci-GA2. We found that single mutations severely reduced the activity of Ci-GA2 (Fig. 5 L and M) and double mutations completely blocked the ability of Ci-GA2 to induce ptc-lacZ expression in P compartment cells (Fig. 5N). Consistently, RNAi of aPKC decreased, whereas overexpression of aPKCCAAX increased, Ci-GA2 activity in a ptc-luc reporter assay (Fig. 5R). Mutating Thr512 alone (Ci-GA2 TA1) attenuated its regulation by aPKC, and mutating both Thr512 and Thr590 (Ci-GA2TA12) further blocked its regulation by aPKC (Fig. 5R). We also found that mutating Thr512 and Thr590 in full-length Ci blocked its regulation by aPKC in the ptc-luc assay (Fig. 5S). These observations suggest that aPKC regulates the activity of Ci in both wing discs and cultured S2 cells, likely through two critical residues: Thr512 and Thr590. We wondered whether aPKC phosphorylates Ci directly, and thus performed an in vitro kinase assay using a purified Ci fragment (amino acids 497–604, cleaved from GST-Ci fusion protein) and recombinant aPKC. We found that aPKC was able to phosphorylate wild type Ci (CiWT) but not mutant Ci carrying the T512, 590A substitution (CiTA12) (Fig. 5P), indicating that aPKC phosphorylates these Thr residues. Because Thr512 and Thr590 fall into the DNA-binding domain and mutating these residues abolishes activity of Ci, we further tested whether phosphorylation promotes the ability of Ci to bind DNA. We carried out an EMSA to examine the interaction between DNA and CiWT or mutant CiDB (amino acids 497–604) after in vitro phosphorylation by PKC. We found that CiDBWT did not bind DNA (Fig. 5Q, lane 2), unless it was phosphorylated by PKC (lane 3). Blocking PKC activity by inhibitor prevented CiDB–DNA binding (Fig. 5Q, lane 4). We further found that CiDBTA12 did not have the ability to bind DNA even though PKC was added to the assay (Fig. 5Q, lanes 5 and 6). Taken together, our data suggest that aPKC positively regulates Ci activity by phosphorylating Ci at Thr512 and Thr590 and promoting Ci to bind DNA. aPKC–Par6 Protein Complex Regulates Hh Signaling in Vivo. It is also possible that the polarity protein complexes regulate Hh signaling through regulating cell polarity. It has been shown that aPKC forms a functional complex with Par6 to control epithelial polarization (12). Indeed, the interaction between the endogenous aPKC and Par6 was readily detected in cultured S2 cells (Fig. S2A). Knockdown of aPKC by RNAi decreased, whereas overexpression of aPKC increased, the level of Par6 (Fig. S2B). Additionally, in wing imaginal discs, aPKC RNAi decreased the level of Par6 and Par6 RNAi decreased the level of aPKC (Fig. S2 C and D). These data suggest that aPKC and Par6 may function as a stable complex to regulate Smo. Strikingly, basal accumulation of Smo in wing disc was blocked in cells with mutated par6 (Fig. 6A). Similarly, knockdown of Par6 by RNAi attenuated Smo accumulation and down-regulated ptc-lacZ expression (Fig. 6 B and C). Overexpressing Par6 alone in wing disc neither caused Smo accumulation nor induced Ci elevation (Fig. 6D). However, coexpressing Par6 with aPKCCAAX induced peak activation of Hh signaling, leading the A compartment cells to occupy the entire Jiang et al.

PNAS PLUS

wing disc (Fig. 6E) and inducing overgrowth of the wing (Fig. S2E), whereas expressing aPKCCAAX alone induced a much lower level of ectopic Hh activity (Fig. 1G). These findings suggest that aPKC and Par6 act together to regulate Hh signaling. If the aPKC–Par6 protein complex regulates Hh signaling through controlling epithelial polarization, we reasoned that other polarity proteins necessary for the aPKC–Par6 complex should also be required. However, inactivation of Baz by either mutation or RNAi did not regulate Smo and Ci accumulation in wing discs (Fig. S3 A and B). Similarly, inactivation of Crb, Dlg5, Dlg1, and Lgl by RNAi did not cause any changes in Smo and Ci accumulation (Fig. S3 D, E, and G). Overexpression of either Baz or Dlg5 had no effect on Hh signaling (Fig. S3C). These findings suggest that aPKC and Par6 are directly involved in Hh signaling, which may not require the other polarity proteins.

Jiang et al.

PNAS | Published online October 27, 2014 | E4847

CHEMISTRY

Fig. 5. Additional role of aPKC in Hh signaling by regulating the Zn finger DNA-binding domain of Ci. (A and B) Wing discs expressing UAS-Myc-Smo680-3SD alone or together with UAS-aPKC-RNAi105624 by MS1096 Gal4 were immunostained for dpp-lacZ and Ci. Arrows indicate the reduced expression of dpp-lacZ and Ci by knockdown of aPKC. (C) WT wing disc was stained for ptc-lacZ. (D and E) Wing discs expressing UAS-Myc-SmoDDD680-3SD alone or together with UAS-aPKC-RNAi105624 by MS1096 Gal4 were immunostained for ptc-lacZ. The arrow in E indicates the reduced expression of ptc-lacZ by aPKC RNAi. (F and G) Wing discs expressing UAS-HA-Ci-3P alone or together with UAS-aPKC-RNAi105624 by MS1096 Gal4 were immunostained for ptclacZ. The arrow in G indicates the reduced expression of ptc-lacZ caused by aPKC RNAi. (H) Wing disc expressing UAS-Ci-GA2 by MS1096 Gal4 was

stained for ptc-lacZ. (I–K) Wing discs coexpressing UAS-Ci-GA2 with UASaPKC RNAi, UAS-aPKCCAAX, or UAS-MC* by MS1096 Gal4 were stained for ptc-lacZ expression. Arrows indicate ptc-lacZ expression induced by Ci-GA2. (L–N) Wing discs expressing the indicated construct by MS1096 Gal4 were stained for ptc-lacZ. Arrows indicate the blockade of ptc-lacZ expression in P compartment cells. (O) Alignment of the Zn finger DNA-binding domain from Ci and Gli proteins. The identical sequences of amino acids around Thr512 and Thr590 are shown. (P) Autoradiograph of the in vitro kinase assays to examine Ci phosphorylation by PKC. The indicated Ci proteins were incubated with recombinant PKCζ in the presence of [γ-32P]ATP (lanes 2–4). PKC inhibitor is used to block the activity of PKCζ (lane 4). The same amount of Ci fragment was used for each reaction (Experimental Procedures), and the experiment was repeated three times. (Q) EMSA assay to examine CiDNA binding. The indicated Ci proteins were incubated with [γ-32P]ATP-labeled oligos that contain the Ci/Gli-binding nucleotides, followed by native polyacrylamide gel separation and autoradiography. In this assay, Ci proteins containing amino acids 497–604 were bacterially expressed, dephosphorylated with calf intestinal alkaline phosphatase, and cleaved with thrombin (Experimental Procedures). The same amount of Ci fragment was used, and the experiment was repeated three times. Control, CiWT with unlabeled oligos. (R) Activity of each Ci-GA2 construct containing the Zn finger DNAbinding domain fused to the Gal4 activation domain was examined using the ptc-luc reporter assay. (S) Activity of the indicated Ci full-length constructs was examined using the ptc-luc reporter assay. The y axis represents normalized ptc-luc activity.

DEVELOPMENTAL BIOLOGY

Hh Signaling Up-Regulates the Expression of aPKC. The elevated immunostaining of aPKC in ptc mutant cells suggests that Hh signaling may regulate aPKC expression (Fig. 3D). The Drosophila wing/wing imaginal disc has been a classic organ used to dissect the mechanisms of Hh signal transduction. The transcription factor Ci is only expressed in the A compartment Hh receiving cells, so the behavior of cells in the A compartment near the A/P boundary has been well characterized. When we examined the expression pattern of endogenous aPKC in wing disc, we found that aPKC levels had mild elevation in A compartment cells near the A/P boundary where Hh receiving cells resided (Fig. 7A), suggesting that Hh may elevate the levels of aPKC in the wing. To verify this notion further, we expressed Hh with the MS1096 Gal4 in wing discs and found elevation of aPKC in A compartment cells away from the A/P boundary, indicating the stabilization of aPKC by Hh expression (Fig. 7B). Furthermore, the elevation of aPKC is likely Ci-dependent, because immunostaining of aPKC was only elevated in A compartment cells that express Ci. In support of this hypothesis, we found that the expression of aPKC was increased in both A and P compartment cells when Ci−3P was expressed (Fig. 7C), and that the expression of aPKC was only elevated in A compartment cells when SmoDDD was expressed (Fig. 7D). Our findings indicate that Hh induces the expression of aPKC, which phosphorylates and activates Smo and Ci, suggesting a positive feedback regulation of Hh signaling components by aPKC (Fig. 7E).

cluster, according to the zipper-lock model we previously described (26), which may account, at least in part, for the ability of aPKC to promote Smo activation. This finding can also explain

Fig. 6. Par6 regulates Smo accumulation and activity. (A) A par6Δ226 mutant clone in the wing disc was stained for Smo. The arrow indicates that Smo accumulation in P compartment cells is blocked by par6 mutation. GFP marks the clone. (B and C) Wing discs expressing UAS-Par6-RNAi108560 by MS1096 Gal4 were stained for Smo, ptc-lacZ, and Ci. Arrows indicate the reduction of Smo and ptc-lacZ by knockdown of Par6. (D and E) Wing discs expressing UAS-Par6 alone or together with UAS-aPKCCAAX by MS1096 Gal4 were stained for Smo, Par6, and Ci. Coexpression of Par6 and aPKC accumulates high levels of Smo and Ci in wing disc.

Discussion It has been shown that Hh protein can travel both apically and basally in cells near the A/P boundary but can only accumulate apically in cells away from the A/P boundary (42), which may guide the accumulation of Smo in apical and basolateral domains in Drosophila wing disc (43). However, it is unknown how Smo accumulation is controlled within the apical and basal domains of the wing and whether apical/basal polarity proteins are involved in the signal transduction. This study identifies aPKC as a positive regulator for Hh signal transduction by regulating both Smo and Ci. Importantly, we discover that aPKC has direct roles in Hh signaling by phosphorylating multiple Hh signaling components. Smo phosphorylation by aPKC at Ser680 promotes Smo activation. aPKC promotes the DNA-binding activity of Ci Zn finger DNA-binding domain through phosphorylating Thr512 and Thr590. These findings indicate that aPKC is directly involved in Hh signaling. We found that phosphodeficient mutation in the second cluster of PKA/CK1 sites blocked the ectopic Smo activity (Fig. 4 J and K) even if the aPKC site (amino acid 680) and adjacent CK1 site (amino acid 683) are mutated to Asp to mimic phosphorylation. It is possible that the main role of phosphorylation by aPKC at S680 and subsequent phosphorylation by CK1 is to promote PKA/CK1mediated phosphorylation at the second cluster, or even the third E4848 | www.pnas.org/cgi/doi/10.1073/pnas.1417147111

Fig. 7. Hh signaling up-regulates aPKC level in the wing. (A) WT wing disc was stained for Ci and PKC. (Right) High-magnification image is shown. Arrows indicate the mild elevation of aPKC in A compartment cells near the A/P boundary. (Middle) Dashed line indicates the A/P boundary that is defined by Ci expression. (B) Wing disc expressing UAS-Hh by MS1096 Gal4 was stained for PKC and Ci. The arrow indicates the elevation of PKC in A compartment cells. (C and D) Wing discs expressing UAS-Ci-3P or UAS-SmoDDD by the dorsal compartment-specific ap-Gal4 were stained for PKC and HA (for HA-Ci-3P) or Myc (for Myc-SmoDDD). The arrows indicates the increased PKC staining in A compartment cells, and the arrowhead indicates the increased PKC staining in P compartment cells. (E) Model of Hh signal transduction that is regulated by aPKC. aPKC directly regulates both Smo and Ci, and Hh signaling up-regulates aPKC expression that is dependent on Ci.

Jiang et al.

Experimental Procedures Constructs, Mutants, and Transgenes. Myc-SmoWT, Myc-SmoDDD, Myc-SmoDAD, and Myc-SmoDSD constructs and the generation of transgenes at the 75B1 attP locus (resulting in VK5 lines) by PhiC31 integration have been described previously (26, 47). Smo variants with mutations in the aPKC and CK1 sites were generated by site-directed mutagenesis, and their transgenes were generated by the same approach. HA-Ci-GA2TA1, HA-Ci-GA2TA2, HA-CiGA2TA12, and Myc-CiTA12 were generated by a similar approach. HA-Par6 and HA-aPKCΔN were constructed by fusion of the Par6 coding sequence and aPKC amino acids 180–606 to the attB-UAST-2HA backbone, respectively. The UAST-aPKC construct was a gift from Jurgen Knoblich (Institute of Molecular Biotechnology, Vienna, Austria). GST-Smo680SA and GST-Smo683SA contained Smo amino acids 656–755 with either the PKC site or the adjacent CK1 site mutated to Ala. After the generation of site mutations, all of the constructs used in this study were confirmed by sequencing. The aPKC RNAi lines were obtained from either the BSC (35001) or VDRC (v105624). Par6

Jiang et al.

PNAS PLUS

RNAi lines were obtained from the VDRC (v108560, v19730, and v19731) and BSC (35000), and line v108560 was used for most of the experiments because all those lines gave rise to similar phenotypes. RNAi lines targeting Ptc (28795), Baz [v105989 and RNAi lines from Denise Montell (Johns Hopkins University School of Medicine, Baltimore, MD)], Crb (27697), and Lgl (31089, 35773, v51247, and v109604) were obtained from the BSC, VDRC, or published sources. Ruth Steward (Rutgers University, Piscataway, NJ) kindly provided Dlg RNAi and overexpression lines. Tony Harris (University of Toronto, Toronto, Canada) provided the bazXi106 mutant strain. MS1096 Gal4, actin>CD2>Gal4, and ap-Gal4 have been described (28, 48). HA-Ci-3P and CiGA2 (39, 41), MC* and sggD127 (36), par6Δ226 and Baz RNAi strains (49), and aPKCK06403 have been described (35). The genotypes used for generating mutant clones are as follows: smo clones expressing UAS-aPKCΔN by the MS1096 Gal4: yw hsp-flp, MS1096/+ or Y; smo3 FRT40/hs-GFP FRT40; UASaPKCΔN. ptc clones: yw hsp-flp/+ or Y; ptc[wII] FRT42D/hs-GFP FRT42D. par6Δ226 MARCM clones: hsp-flp MARCM FRT19A tub-Gal80 UAS-GFP/ par6Δ226 FRT19. aPKC clones: yw hsp-flp, MS1096/+ or Y; aPKCK06403 FRT42D/hs-GFP FRT42D; UAS-P35. sgg clones: hsp-flp sggD127/Y; FRT82B hsCD2, y+Dp(1;3)w+67K, sgg+/FRT82B hs-Myc.

In Vitro Kinase Assay, EMSA Assay, and Immunostaining of Wing Imaginal Discs. For the in vitro kinase assay, GST-Smo fusion proteins were expressed in bacteria, purified with GST beads, and cleaved with thrombin according to the manufacturer’s instructions (GE Healthcare Life Science). Three micrograms of Smo was incubated at 30 °C for 30 min in 50 μL of assay buffer containing 35 mM Tris (pH 7.5), 10 mM MgCl2, 0.1 mM CaCl2, 0.5 mM EGTA, and 2.5 μM ATP or 10 μCi of γ-[32P]ATP in the presence of commercial recombinant human PKCζ (CD BioSciences) and CK1 (New England Biolabs), followed by autoradiography. Inhibitors used are as follows: myristoylated PKCζ pseudosubstrate inhibitor (10 μM, catalog no. 539624; Millipore) and CK1–7 dihydrochloride (10 μM, catalog no. 1177141-67-1; Sigma–Aldrich) (22). GST-CiWT and GST-CiTA12 containing Ci amino acids 497–604 with either WT or Thr512/Thr590 mutated to Ala were expressed in bacteria, purified with GST beads, cleaved with thrombin, and subjected to the in vitro kinase assay with the same methods as used for Smo. To ensure an equal amount of Ci fragment in each reaction, the same amount of GST-Ci protein was used to achieve Ci fragmentation, with the same conditions for cleavage. For the EMSA DNA-binding assay, GST-Ci proteins were bacterially expressed; purified using the method described above; dephosphorylated with calf intestinal alkaline phosphatase (catalog no. 10713023001; Roche) in 50 μL of buffer containing 50 mM Tris·HCl (pH 7.9), 10 mM MgCl2, 100 mM NaCl2, 1 mM DTT, and protease inhibitor at 37 °C for 60 min; and finally cleaved with thrombin. The DNA probe containing the Ci/Gli-binding site was generated by

PNAS | Published online October 27, 2014 | E4849

DEVELOPMENTAL BIOLOGY

Cell Culture, Transfection, Western Blot, and Luciferase Reporter Assay. S2 cells were cultured as previously described (26). Briefly, transfections were carried out using Effectene transfection reagent (Qiagen). Forty-eight hours posttransfection, cells were harvested and treated with lysis buffer [100 mM NaCl, 50 mM Tris·HCl (pH 8.0), 1.5 mM EDTA, 10% (vol/vol) glycerol, 1% (vol/vol) Nonidet P-40, and protease inhibitor tablet (Roche)]. Cell lysate was obtained by centrifuging at 13,200 × g for 10 min. A total of 6 × 106 cells were harvested and lysed in 450 μL of lysate buffer. Fifty microliters was saved for direct Western blots, of which 4 μL was used for each load. The remaining 400 μL was used for immunoprecipitation assay, in which cell lysate was added with beads of Protein A Ultralink Resin (Thermo Scientific) after adding the proper primary antibody for 2 h. The samples were then resolved by SDS/PAGE and transferred onto PVDF membranes (Millipore) for Western blotting. Western blot analysis was performed using the indicated antibodies and the ECL protocol. The use of Hh-conditioned medium has been described previously (26). Treating S2 cells with OA or dsRNA has been described previously (26, 47). The aPKC dsRNA was synthesized against ORF nucleotides 379–930. Antibodies used for Western blotting are as follows: mouse anti-HA (F7, 1:5,000; Santa Cruz Biotechnology), anti–β-tubulin (concentrated, 1:5,000; Developmental Studies Hybridoma Bank); rabbit anti-PKC (sc-216, 1:10,000; Santa Cruz Biotechnology), and anti-Par6 (1:5,000; a gift from Denise Montell). The ptc-luc reporter assay has been described with S2 cells transfected with tub-Ci (26). The Phos-tag gel assay was carried out as described previously (33). Briefly, 20 μM Phos-tag acrylamide (AAL-107; Wako Chemical) and 40 μM MnCl2 were included. Before electroblotting, the gels were soaked in transfer buffer containing 2 mM EDTA for 15 min to eliminate the manganese ion, followed by soaking in transfer buffer without EDTA for 15 min with gentle agitation. The duration of electroblotting was extended to 3 h. Densitometric analysis of Myc-Smo bands on the Phos-tag gels was done using ImageJ software (NIH, version 1.48).

CHEMISTRY

why S680 phosphorylation is dispensable in SmoDDD background. It would be interesting to test this hypothesis in the future using phosphospecific antibodies to examine PKA/CK1-mediated phosphorylation at these individual sites in the presence or absence of aPKC-mediated phosphorylation at S680. Both the loss and gain of aPKC studies indicate that aPKC promotes basolateral accumulation of Smo, which correlates with Hh pathway activation. The aPKC may regulate Smo subcellular localization by directly phosphorylating Smo at S680 and/or through other mechanisms. The examination of several polarity proteins, including Crb, Baz, Dlg5, Dlg1, and Lgl, showed that inactivation of these proteins produced polarity phenotypes but did not affect Smo accumulation and Hh signal transduction in Drosophila wings (Fig. S3), suggesting that aPKC regulates the activity of Smo and Ci independent of its function in apical/basal polarity. However, we did observe that Smo accumulation and Hh signaling were affected by loss of Par6 function. Furthermore, coexpression of Par6 synergized with aPKCCAAX to promote Smo accumulation and Hh pathway activation. We found that Par6 and aPKC form a complex and that Par6 is required for aPKC stabilization, which may explain why Hh signaling is affected by loss of Par6. In addition to regulating aPKC stability, Par6 may facilitate aPKC to phosphorylate Smo and Ci, because an adaptor function of Par6 has been identified (44, 45). Another possibility is that Par6 activates aPKC (46), which, in turn, promotes Smo phosphorylation. It is also possible that Par6 has an additional function beyond coupling aPKC in regulating Smo activation because mutating par6 caused a more severe Smo reduction (Fig. 6A) compared with the inactivation of aPKC (Fig. 2B). Future study is needed to explore the precise mechanism by which the aPKC– Par6 protein complex regulates Smo accumulation and Hh signal transduction. Abnormal Smo activation has been found in BCC and medulloblastoma. In mammals, Hh induces Smo ciliary localization, and it has been speculated that dysregulation of Smo localization in the cilium will cause cancer. A recent study indicates that aPKC is accumulated in the cilia of BCC cells (40). In addition, this study revealed that aPKC phosphorylates Gli1 to regulate its DNA-binding activity. Our study thus uncovers a conserved role of aPKC in the regulation of Ci/Gli DNA-binding activity. In addition, we find that the aPKC–Par6 complex regulates Smo subcellular localization. Given the interesting parallel between membrane accumulation of Smo in Drosophila and translocation of Smo to cilium in mammalian cells, it is possible that malfunction of aPKC will lead to aberrant Smo ciliary localization and activation, and thus cancer formation. Further investigation of whether Smo ciliary localization is regulated by aPKC will provide mechanistic insight into Smo-dependent tumorigenesis. The direct roles of the aPKC–Par6 complex in regulating Hh signaling components discovered in this study also raise the possibility that mammalian aPKC could regulate Hh signaling during embryonic development.

annealing two oligos and labeled with γ-[32P]ATP by Klenow fragment (Roche). The oligos were CiBS1 (5′-CACTGAGAT TGGGTGGTC TATCAGTT-3′) and CiBS2 (5′-AACTGATA GACCACCCA ATCTCAGTG-3′), with the 9-bp core consensus Ci/Gli-binding site underlined (50). For each reaction, 20 μL of incubation buffer was used, which contained 15 mM Hepes (pH 7.9), 80 mM NaCl, 15 mM KCl, 20 μM EDTA, 1 mM DTT, 3% glycerol, and 1 μM probe. PKCζ and PKCζ pseudosubstrate inhibitor were added if necessary. Samples were separated with 8% (vol/vol) native polyacrylamide gel in TBE (Tris/ Borate/EDTA, a buffer solution containing a mixture of Tris base, boric acid and EDTA), followed by autoradiography. For imaginal disc immunostaining, wing discs from third instar larvae with specific genotypes were dissected in PBS and then fixed with 4% (vol/vol) formaldehyde in PBS for 20 min. After permeabilization with PBT (PBS supplemented with 1% Triton X100), discs were incubated with the indicated primary antibodies for 3 h and the corresponding secondary antibodies for 1 h sequentially, and then washed with PBT three times, for 20 min per wash, following incubations. Antibodies used in this study are as follows: mouse anti-SmoN (1:10; Developmental Studies Hybridoma Bank), anti-Ptc (1:10;

ACKNOWLEDGMENTS. We thank Drs. Xuesong Cao and Hongge Jia for help with Ci-GA2 and GST-Ci constructs and transgenic lines. We thank Dr. Gregory Longmore for the par6Δ226 mutant; Dr. Wu-Min Deng for the aPKCK06403 mutant; Dr. Joan Hooper for the ptc mutant and UAS-P35 flies; Dr. Juergen Knoblich for the aPKC construct and UAS-aPKCΔN and UAS-aPKCCAAX flies; Dr. Denise Montell for the anti-Par6 antibody, par6Δ226 mutant, Baz RNAi, and UAS-Par6 fly lines; and Dr. Tony Harris for the bazXi106 stock. We thank Dr. Ruth Steward for sharing unpublished findings and the Dlg RNAi and overexpression lines. We thank the Developmental Studies Hybridoma Bank for antibodies and the BSC and VDRC for fly stocks. J. Jia is supported by NIH Grant GM079684, T. Gao is supported by NIH Grant CA133429, and J. Jiang is supported by NIH Grants GM061269 and GM067405 and by Grant I-1603 from the Welch Foundation.

1. Nüsslein-Volhard C, Wieschaus E (1980) Mutations affecting segment number and polarity in Drosophila. Nature 287(5785):795–801. 2. Ingham PW, Nakano Y, Seger C (2011) Mechanisms and functions of Hedgehog signalling across the metazoa. Nat Rev Genet 12(6):393–406. 3. Jiang J, Hui CC (2008) Hedgehog signaling in development and cancer. Dev Cell 15(6): 801–812. 4. Ingham PW, McMahon AP (2001) Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 15(23):3059–3087. 5. Basler K, Struhl G (1994) Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein. Nature 368(6468):208–214. 6. Tabata T, Kornberg TB (1994) Hedgehog is a signaling protein with a key role in patterning Drosophila imaginal discs. Cell 76(1):89–102. 7. Lecuit T, et al. (1996) Two distinct mechanisms for long-range patterning by Decapentaplegic in the Drosophila wing. Nature 381(6581):387–393. 8. Nellen D, Burke R, Struhl G, Basler K (1996) Direct and long-range action of a DPP morphogen gradient. Cell 85(3):357–368. 9. Campbell G, Tomlinson A (1999) Transducing the Dpp morphogen gradient in the wing of Drosophila: Regulation of Dpp targets by brinker. Cell 96(4):553–562. 10. Jia J, Jiang J (2006) Decoding the Hedgehog signal in animal development. Cell Mol Life Sci 63(11):1249–1265. 11. Hooper JE, Scott MP (2005) Communicating with Hedgehogs. Nat Rev Mol Cell Biol 6(4):306–317. 12. Goldstein B, Macara IG (2007) The PAR proteins: Fundamental players in animal cell polarization. Dev Cell 13(5):609–622. 13. Tepass U, Theres C, Knust E (1990) crumbs encodes an EGF-like protein expressed on apical membranes of Drosophila epithelial cells and required for organization of epithelia. Cell 61(5):787–799. 14. Lemmers C, et al. (2004) CRB3 binds directly to Par6 and regulates the morphogenesis of the tight junctions in mammalian epithelial cells. Mol Biol Cell 15(3):1324–1333. 15. Wodarz A, Hinz U, Engelbert M, Knust E (1995) Expression of crumbs confers apical character on plasma membrane domains of ectodermal epithelia of Drosophila. Cell 82(1):67–76. 16. Bilder D, Schober M, Perrimon N (2003) Integrated activity of PDZ protein complexes regulates epithelial polarity. Nat Cell Biol 5(1):53–58. 17. Tanentzapf G, Tepass U (2003) Interactions between the crumbs, lethal giant larvae and bazooka pathways in epithelial polarization. Nat Cell Biol 5(1):46–52. 18. Zheng X, Mann RK, Sever N, Beachy PA (2010) Genetic and biochemical definition of the Hedgehog receptor. Genes Dev 24(1):57–71. 19. Lum L, Beachy PA (2004) The Hedgehog response network: Sensors, switches, and routers. Science 304(5678):1755–1759. 20. Zhang C, Williams EH, Guo Y, Lum L, Beachy PA (2004) Extensive phosphorylation of Smoothened in Hedgehog pathway activation. Proc Natl Acad Sci USA 101(52): 17900–17907. 21. Apionishev S, Katanayeva NM, Marks SA, Kalderon D, Tomlinson A (2005) Drosophila Smoothened phosphorylation sites essential for Hedgehog signal transduction. Nat Cell Biol 7(1):86–92. 22. Jia J, Tong C, Wang B, Luo L, Jiang J (2004) Hedgehog signalling activity of Smoothened requires phosphorylation by protein kinase A and casein kinase I. Nature 432(7020):1045–1050. 23. Zhao Y, Tong C, Jiang J (2007) Hedgehog regulates smoothened activity by inducing a conformational switch. Nature 450(7167):252–258. 24. Jia H, et al. (2010) Casein kinase 2 promotes Hedgehog signaling by regulating both smoothened and Cubitus interruptus. J Biol Chem 285(48):37218–37226. 25. Chen Y, et al. (2010) G protein-coupled receptor kinase 2 promotes high-level Hedgehog signaling by regulating the active state of Smo through kinase-dependent and kinase-independent mechanisms in Drosophila. Genes Dev 24(18):2054–2067. 26. Fan J, Liu Y, Jia J (2012) Hh-induced Smoothened conformational switch is mediated by differential phosphorylation at its C-terminal tail in a dose- and positiondependent manner. Dev Biol 366(2):172–184.

27. Yang X, et al. (2013) Drosophila Vps36 regulates Smo trafficking in Hedgehog signaling. J Cell Sci 126(Pt 18):4230–4238. 28. Xia R, Jia H, Fan J, Liu Y, Jia J (2012) USP8 promotes smoothened signaling by preventing its ubiquitination and changing its subcellular localization. PLoS Biol 10(1): e1001238. 29. Rohatgi R, Milenkovic L, Scott MP (2007) Patched1 regulates hedgehog signaling at the primary cilium. Science 317(5836):372–376. 30. Rohatgi R, Milenkovic L, Corcoran RB, Scott MP (2009) Hedgehog signal transduction by Smoothened: Pharmacologic evidence for a 2-step activation process. Proc Natl Acad Sci USA 106(9):3196–3201. 31. Wang Y, Zhou Z, Walsh CT, McMahon AP (2009) Selective translocation of intracellular Smoothened to the primary cilium in response to Hedgehog pathway modulation. Proc Natl Acad Sci USA 106(8):2623–2628. 32. Corbit KC, et al. (2005) Vertebrate Smoothened functions at the primary cilium. Nature 437(7061):1018–1021. 33. Chen Y, et al. (2011) Sonic Hedgehog dependent phosphorylation by CK1α and GRK2 is required for ciliary accumulation and activation of smoothened. PLoS Biol 9(6): e1001083. 34. Wilson CW, Chuang PT (2010) Mechanism and evolution of cytosolic Hedgehog signal transduction. Development 137(13):2079–2094. 35. Tian AG, Deng WM (2008) Lgl and its phosphorylation by aPKC regulate oocyte polarity formation in Drosophila. Development 135(3):463–471. 36. Jia J, et al. (2002) Shaggy/GSK3 antagonizes Hedgehog signalling by regulating Cubitus interruptus. Nature 416(6880):548–552. 37. Price MA, Kalderon D (2002) Proteolysis of the Hedgehog signaling effector Cubitus interruptus requires phosphorylation by Glycogen Synthase Kinase 3 and Casein Kinase 1. Cell 108(6):823–835. 38. Kinoshita E, Kinoshita-Kikuta E, Takiyama K, Koike T (2006) Phosphate-binding tag, a new tool to visualize phosphorylated proteins. Mol Cell Proteomics 5(4):749–757. 39. Wang G, Wang B, Jiang J (1999) Protein kinase A antagonizes Hedgehog signaling by regulating both the activator and repressor forms of Cubitus interruptus. Genes Dev 13(21):2828–2837. 40. Atwood SX, Li M, Lee A, Tang JY, Oro AE (2013) GLI activation by atypical protein kinase C ι/λ regulates the growth of basal cell carcinomas. Nature 494(7438):484–488. 41. Wang G, Amanai K, Wang B, Jiang J (2000) Interactions with Costal2 and suppressor of fused regulate nuclear translocation and activity of cubitus interruptus. Genes Dev 14(22):2893–2905. 42. Ayers KL, Gallet A, Staccini-Lavenant L, Thérond PP (2010) The long-range activity of Hedgehog is regulated in the apical extracellular space by the glypican Dally and the hydrolase Notum. Dev Cell 18(4):605–620. 43. Denef N, Neubüser D, Perez L, Cohen SM (2000) Hedgehog induces opposite changes in turnover and subcellular localization of patched and smoothened. Cell 102(4): 521–531. 44. Joberty G, Petersen C, Gao L, Macara IG (2000) The cell-polarity protein Par6 links Par3 and atypical protein kinase C to Cdc42. Nat Cell Biol 2(8):531–539. 45. Etienne-Manneville S, Hall A (2003) Cdc42 regulates GSK-3beta and adenomatous polyposis coli to control cell polarity. Nature 421(6924):753–756. 46. Graybill C, Wee B, Atwood SX, Prehoda KE (2012) Partitioning-defective protein 6 (Par-6) activates atypical protein kinase C (aPKC) by pseudosubstrate displacement. J Biol Chem 287(25):21003–21011. 47. Jia H, Liu Y, Yan W, Jia J (2009) PP4 and PP2A regulate Hedgehog signaling by controlling Smo and Ci phosphorylation. Development 136(2):307–316. 48. Jia J, Tong C, Jiang J (2003) Smoothened transduces Hedgehog signal by physically interacting with Costal2/Fused complex through its C-terminal tail. Genes Dev 17(21): 2709–2720. 49. Pinheiro EM, Montell DJ (2004) Requirement for Par-6 and Bazooka in Drosophila border cell migration. Development 131(21):5243–5251. 50. Von Ohlen T, Lessing D, Nusse R, Hooper JE (1997) Hedgehog signaling regulates transcription through cubitus interruptus, a sequence-specific DNA binding protein. Proc Natl Acad Sci USA 94(6):2404–2409.

E4850 | www.pnas.org/cgi/doi/10.1073/pnas.1417147111

Developmental Studies Hybridoma Bank), anti-Myc (9E10, 1:50; Santa Cruz Biotechnology), anti-En (1:20; Developmental Studies Hybridoma Bank), rabbit anti-PKC (sc-216, 1:1,000; Santa Cruz Biotechnology), anti-Par6 (1:500; a gift from Denise Montell), anti–β-gal (1:1,500; Cappel), and rat anti-Ci (2A, 1:10; Developmental Studies Hybridoma Bank).

Jiang et al.