Ptch2 shares overlapping functions with Ptch1 in Smo ... shares overlapping functions with Ptch1 in Smo ... including exencephaly, open neural tube and ... and regulating Smo through

Ptch2 shares overlapping functions with Ptch1 in Smo regulationand limb development

Olena Zhulyn a,b, Erica Nieuwenhuis a, Yulu Cherry Liu c,Stephane Angers c, Chi-chung Hui a,b,n

a Program in Developmental & Stem Cell Biology, The Hospital for Sick Children, Toronto, Ontario, Canada M5G 1X8b Department of Molecular Genetics, University of Toronto, Toronto, Ontario, Canada M5S 1A8c Department of Biochemistry and Pharmaceutical Sciences, University of Toronto, Ontario, Canada M5S 1A8

a r t i c l e i n f o

Article history:Received 22 May 2014Received in revised form4 October 2014Accepted 27 October 2014Available online 13 November 2014

Keywords:SmoothenedPatchedPatterningSonic hedgehogApical ectodermal ridgeMesenchyme

a b s t r a c t

Ptch1 and Ptch2 are highly conserved vertebrate homologs of Drosophila ptc, the receptor of theHedgehog (Hh) signaling pathway. The vertebrate Ptch1 gene encodes a potent tumor suppressor and iswell established for its role in embryonic development. In contrast, Ptch2 is poorly characterized anddispensable for embryogenesis. In flies and mice, ptc/Ptch1 controls Hh signaling through the regulationof Smoothened (Smo). In addition, Hh pathway activation also up-regulates ptc/Ptch1 expression torestrict the diffusion of the ligand. Recent studies have implicated Ptch2 in this ligand dependentantagonism, however whether Ptch2 encodes a functional Shh receptor remains unclear. In this report,we demonstrate that Ptch2 is a functional Shh receptor, which regulates Smo localization and activityin vitro. We also show that Ptch1 and Ptch2 are co-expressed in the developing mouse limb bud and lossof Ptch2 exacerbates the outgrowth defect in the limb-specific Ptch1 knockout mutants, demonstratingthat Ptch1 and Ptch2 co-operate in regulating cellular responses to Shh in vivo.

& 2014 Elsevier Inc. All rights reserved.

Introduction

Sonic hedgehog (Shh) is an important regulator of patterning,development and homeostasis in the embryo and adult (Hui andAngers, 2011; Jiang and Hui, 2008). The Shh signal is transduced bythe transmembrane protein Patched 1 (Ptch1) through regulationof Smoothened (Smo) (Chen and Struhl, 1996; Hooper and Scott,1989; Ingham et al., 2000). In the absence of Shh, Ptch1 inhibitsSmo activity. Binding of Shh to Ptch1 alleviates this inhibitionresulting in Smo activation. The precise mechanism of Smoactivation is not known, but it requires the translocation of Smofrom the plasma membrane into the primary cilium (Corbit et al.,2005; Dorn et al., 2012; Milenkovic et al., 2009; Rana et al., 2013;Rohatgi et al., 2007, 2009; Wang et al., 2009; Wilson et al., 2009).Accumulation of Smo in the primary cilium promotes the activa-tion of transcription factors Gli1, Gli2 and Gli3, the effectors of Shhsignaling, and Gli target gene expression (Hui and Angers, 2011).

Ptch1 regulates Shh signaling through two distinct mechanisms– ligand dependent antagonism (LDA) and ligand independentantagonism (LIA) – which are conserved from flies to mice

(Briscoe et al., 2001; Chen and Struhl, 1996; Goodrich et al.,1996; Holtz et al., 2013). LIA refers to the ability of Ptch1 toconstitutively inhibit Smo in the absence of Shh. In contrast, LDAinvolves the transcriptional up-regulation of Ptch1 mRNA and theaccumulation of Ptch1 protein at the cell surface in response toShh. It is believed that LDA serves to restrict the diffusion range ofthe Shh ligand, thereby regulating the Shh gradient required forpatterning (Briscoe et al., 2001; Chuang and Mcmahon, 1999;Holtz et al., 2013; Jeong and McMahon, 2005).

Elegant studies in the mouse neural tube demonstrated thatthe Shh-target gene Hip (Hedgehog interacting protein) alsoaccumulates at the cell surface in response to pathway activationand co-operates with Ptch1 to restrict the diffusion of Shh (Chuangand Mcmahon, 1999; Holtz et al., 2013; Jeong and McMahon,2005). This confirmed that LDA is critical for establishment ofneuronal cell fate which is specified by discrete levels of Shhpathway activity along the dorsoventral axis (Briscoe et al., 2001;Chuang and Mcmahon, 1999; Jeong and McMahon, 2005).

Recent studies in the chick limb revealed that diffusion ofShh is dependent on transport along stabilized filopodia anddirect transfer, from signal secreting to signal receiving cells. Themembrane glycoproteins Cdo and Boc are required for Shh signaltransduction and were shown to intercept the ligand on signalreceiving cells in the limb (Allen et al., 2011; Izzi et al., 2011;

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/developmentalbiology

Developmental Biology

http://dx.doi.org/10.1016/j.ydbio.2014.10.0230012-1606/& 2014 Elsevier Inc. All rights reserved.

n Corresponding author at: Program in Developmental & Stem Cell Biology,The Hospital for Sick Children, Toronto, Ontario, Canada M5G 1X8.

E-mail address: [email protected] (C.-c. Hui).

Developmental Biology 397 (2015) 191–202

Kavran et al., 2010; Sanders et al., 2013). Cdo and Boc show bothoverlapping and distinct expression patterns in Hh-responsivetissues and their loss has tissue-specific effects, which partlyrecapitulate the phenotype of the Shh� /� mutants (Chiang et al.,2001; Kraus et al., 2001; Okada et al., 2006; Tenzen et al., 2006;Zhang et al., 2006, 2011). Similarly, Gas1 is highly expressed in Hh-responsive tissues of the chick and mouse where it binds andpromotes Hh transduction in signal receiving cells, particularly inregions where ligand concentration is low (Lee and Fan, 2001; Leeet al., 2001; Martinelli and Fan, 2007; Seppala et al., 2007).Notably, Cdo, Boc and Gas1 form distinct complexes with Ptch1and both in vivo and in vitro studies have demonstrated that theseglycoproteins are critical components of the signal transductionmachinery (Allen et al., 2007, 2011; Holtz et al., 2013; Izzi et al.,2011; Seppala et al., 2007).

In addition to these proteins, vertebrates also encode a Ptch1homologue – Ptch2 – which is expressed in many Shh-responsivetissues, including the neural tube, the limb and the skin(Motoyama et al., 1998a, 1998b; Pearse et al., 2001). Until recently,the function of Ptch2 was poorly understood. Ptch2 mutant mice(Ptch2� /� and Ptch2lacZ/lacZ), generated in our laboratory, displayno overt developmental defects, and are viable and fertile(Adolphe et al., 2014; Nieuwenhuis et al., 2006). Similarly, Ptch2mutants independently generated by others are also grosslynormal (Holtz et al., 2013; Lee et al., 2006). This is in stark contrastto Ptch1� /� mutant mice, which are embryonic lethal by mid-gestation (E9.5) and exhibit severe defects consistent with acti-vated Shh signaling, including exencephaly, open neural tube andcardiac defects (Ellis et al., 2003; Goodrich et al., 1997).

Recent work by Holtz et al. (2013) revealed that Ptch2 interactswith Cdo, Boc and Gas1 in vitro and cooperates with Hip and Ptch1to regulate the Shh gradient in the embryonic neural tube throughLDA. It remains unclear whether Ptch2 contributes to LIA asevidence for a role for Ptch2 in Smo regulation and Shh signaltransduction is limited. In particular, over-expression studiesutilizing human PTCH1 and PTCH2 isoforms suggested that whileboth homologs can bind and internalize Shh, only PTCH1 is able toregulate the expression of a Shh-dependent luciferase reporter(Carpenter et al., 1998; Motoyama et al., 1998a; Rahnama et al.,2004). However, work from our laboratory and others has shownthat murine Ptch2 is able to inhibit the activity of Shh/Gli-responsive reporters and that Ptch2 responds to Shh in transfec-tion assays (Holtz et al., 2013; Nieuwenhuis et al., 2006). Further-more, studies of zebrafish somite and fin development as well asmouse skin development and brain tumorigenesis have suggestedthat Ptch1 and Ptch2 play overlapping roles in pathway regulation(Adolphe et al., 2014; Koudijs et al., 2008; Lee et al., 2006). Thus, itis important to establish if Ptch2 functions as a receptor bytransducing the Shh signal and regulating Smo through LIA. Toaddress this question, we performed biochemical and geneticexperiments to determine whether Ptch2 is a functional Shhreceptor in vitro and in vivo. Our results indicate that, in theabsence of Ptch1, Ptch2 plays a critical role in the regulationof Smo at the primary cilium. This LIA function of Ptch2 comple-ments its role in LDA and gradient regulation in Shh signaling(Holtz et al., 2013).

Results and discussion

Ptch2 over-expression reconstitutes normal Shh signaling in Ptch1� /�

MEFs

Although Ptch1 and Ptch2 share 56% identity at the amino acidlevel, it is not clear if their function is biochemically similar(Kawamura et al., 2008; Motoyama et al., 1998a). The two proteins

differ mostly in the C-terminal region, which is truncated in Ptch2(Carpenter et al., 1998). In vitro analysis suggests that both humanPTCH1 and PTCH2 bind to Shh and the Shh co-receptors Gas1, Cdoand Boc (Bae et al., 2011; Carpenter et al., 1998; Holtz et al., 2013;Izzi et al., 2011). However, mutant analysis revealed that theirrequirement in mouse development is drastically different – whilePtch2� /� mutants are viable and fertile, Ptch1� /� mutants arelethal at E9.5 and exhibit severe defects consistent with hyper-activation of Shh signaling (Goodrich et al., 1997; Lee et al., 2006;Nieuwenhuis et al., 2006).

Holtz et al. (2013) recently demonstrated that Ptch2 inhibitsShh signaling using luciferase reporter and chick neural tubeelectroporation assays. However, it is unclear if Ptch1 and Ptch2use the same mechanism to inhibit the pathway and if they worktogether. It is well established that Ptch1� /� mouse embryonicfibroblasts (MEFs) exhibit constitutive pathway activation (Rohatgiet al., 2007; Taipale et al., 2000). Ptch2 is transcriptionallyupregulated in response to Shh and primary limb fibroblasts,derived from Prx1-Cre;Ptch1f/� mutant forelimbs (E12.5), exhibittranscriptional upregulation of Ptch2 (Fig. S1A) suggesting thatthese cells may be responsive to Shh. To address this question, wetreated Ptch1� /� MEFs with recombinant Shh protein or carrier(BSA) and assessed pathway activation. It was previously shownthat loss of Ptch1 results in constitutive Smo localization to theprimary cilium and pathway activation in MEFs (Rohatgi et al.,2007). Consistent with this, we observe Smo localization in morethan 80% of primary cilia in Ptch1� /� MEFs. Treatment with Shh,but not BSA (carrier), results in a 410% increase in the number ofSmoþ primary cilia (Fig. 1A–C', quantified in Fig. 1D). This isassociated with a 1.4-fold increase in Gli1 (Fig. 1E) and 1.6-foldincrease in Hip (Fig. 1F) mRNA expression indicating that Ptch1� /�

MEFs remain sensitive to Shh ligand. Similarly, we demonstratedthat Ptch1-mutant fibroblast cells, derived from limbs of Prx1-Cre;Ptch1f/� mutant E12.5 embryos, remain sensitive to Shh-conditioned media as shown by upregulation of Gli2 (Fig. 1G)and Gli1 mRNA (Fig. 1H). These findings are consistent with therecent work of Alfaro et al. (2014), which demonstrated thatPtch1� /� cells remain sensitive to Shh independent of Ptch1antiporter activity.

Previous studies showed that over-expression of Ptch1 is ableto partly rescue the phenotypes of Ptch1� /� mouse mutants(Milenkovic et al., 1999). To address the functional similaritybetween Ptch1 and Ptch2, and to determine if Ptch2 mediatesSmo localization and Shh responsiveness in Ptch� /� MEFs, wetested if retroviral infection with Ptch1HA and Ptch2myc con-structs, can suppress constitutive pathway activity in Ptch1� /�

MEFs (Fig. 2A–F). Ptch1� /� MEFs were infected with retroviralvectors carrying Ptch1HA-IRESGFP, Ptch2myc-IRESGFP or with anempty vector (encoding IRESGFP) and GFPþ cells were selected byflow cytometry. To test if GFPþ cells expressed Ptch1HA orPtch2myc protein, we performed Western blot analysis on celllines after cell sorting for GFP. Incubation of whole cell lysates withantibody against HA (Fig. S1D) or Ptch1 (Fig. S1E') detectedabundant Ptch1 protein in Ptch1� /�::Ptch1HA cells. Similarly,antibody against c-myc (Fig. S1E) and Ptch2 (Fig. S1D') detectedPtch2 protein in Ptch1� /�::Ptch2myc cells. Similar expression wasobserved in membrane extracts from Ptch1� /�::Ptch1HA andPtch1� /�::Ptch2myc cells (data not shown). Notably, endogenousPtch2 level appears low in untreated Ptch1� /� MEFs (Fig. S1D').Anti-Ptch1 antibody could not detect endogenous Ptch1 inPtch2� /� primary limb fibroblast (control) (Fig. S1E'). This sug-gested that retroviral infection followed by cell sorting successfullyenriched for Ptch1� /� MEFs expressing high levels of exogenousPtch1 and Ptch2 protein. To compare the effects of Ptch1 and Ptch2on Smo regulation, we examined Smo localization. We found that96% of Ptch1� /� MEFs exhibit constitutive localization of Smo to

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202192

primary cilia (Fig. 2A and A', quantified in Fig. 2D). Over-expression of Ptch2myc results in robust (5.4-fold) suppression ofciliary localization (Fig. 2C and C', quantified in Fig. 2D), whichis comparable to the effects of Ptch1HA overexpression(Fig. 2C and C', quantified in Fig. 2D). Retroviral infection wasassociated with inhibition of transcriptional activity (Fig. S1B),however, over-expression of either Ptch1HA or Ptch2myc leads to afurther, significant reduction in the expression of Shh-target genesGli1 and Hip (Fig. 2E and F). This is consistent with the recentstudy by Holtz et al. (2013) and demonstrates that Ptch2 is apotent negative regulator of Shh signaling which acts directlythrough Smo.

To test whether Ptch2myc can fully reconstitute Shh signaling inPtch1� /� MEFs, we treated Ptch1� /�::Ptch2myc cells with

recombinant Shh protein and assessed their responsiveness tothe ligand. Shh treatment effectively reversed Smo inhibition(resulting in a 5.8-fold increase in ciliary localization) (Fig. 2G,G', J–K', quantified in Fig. 2L) and activated the expression ofGli1 (8.1-fold increase) (Fig. 2M) in Ptch1� /� MEFs expressingPtch2myc. This response is comparable to that in Ptch1HAoverexpression (Fig. 2H–I', L and M). Similarly, Gli1 proteinexpression was suppressed in the presence of Ptch1HA orPtch2myc, while Gli3R levels increased (Fig. S1C). Furthermore,pathway inhibition by Ptch2myc or Ptch1HA could be reversedefficiently by treatment with Smoothened Agonist (SAG), a drugwhich activates Smo (Chen et al., 2002) (Figs. S2A–G). Theseresults suggest that Ptch2 is a functional receptor of Shh thatregulates Smo activity.

Fig. 1. Ptch1 mutant MEFs are responsive to Shh: (A–C') treatment with Shh promotes increased Smo (red) localization (A and A') to α-acetylated tubulin positive (green)primary cilia (B and B') in Ptch1� /� MEFs. (C and C') Merged image. (D) 1.16 fold increase in percentage of cells with Smo accumulation in the cilia upon Shh treatment. (E) 1.4fold increase in the level of Gli1 mRNA and (F) 1.6 fold increase in Hip mRNA expression upon Shh treatment of constitutively activated Ptch1� /� MEFs. (G and H) Limbderived Prx1-Cre;Ptch1f/� fibroblasts exhibit increased Gli2 (G) and Gli1 (H) mRNA levels upon treatment with Shh conditioned media indicating that they are also responsiveto further stimulation with ligand. *Statistically significant difference, po0.05, Student's t-test.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202 193

Ptch1 and Ptch2 are expressed in the ectoderm and mesenchymeof the vertebrate limb

Next, we examined the functional overlap of Ptch1 and Ptch2 inthe developing mouse limb. Previously, we demonstrated that loss ofPtch2 does not affect the patterning of the limb (Nieuwenhuis et al.,2006). Similarly, homozygosity for a novel Ptch2 null allele (Ptch2lacZ)does not affect digit patterning or development (Fig. S3A–G)although Ptch2 mRNA is highly expressed in the limb (Marigoet al., 1996; Motoyama et al., 1998b; Pearse et al., 2001). Notably,loss of Ptch2 did not result in significant changes in Ptch1 mRNA (Fig.S3H) and Ptch2� /� primary limb fibroblasts remained responsive toShh (Fig. S3I). We made use of lacZ reporter alleles of Ptch1 and Ptch2to track expression across limb development. We showed that,similar to the chick, the two genes are co-expressed in the Shh-responsive posterior mesenchyme and apical ectodermal ridge (AER)of the limb (Fig. 3A–F) (Bouldin et al., 2010; Pearse et al., 2001;Scherz et al., 2007). However, Ptch1 exhibits an earlier onset and ahigher level of expression, in both the AER and posterior mesench-yme, across all stages compared to Ptch2 (Fig. 3A–F, Figs. S3A, B, D,and E). Cell sorting of lacZþ cell from Ptch2lacZ/þ embryonic limbbuds at E10.5 confirmed that Ptch1 and Ptch2 are co-expressed inShh-responsive cells (Fig. S4G). Ptch1 and Ptch2 are expressed inresponse to Shh in the neural tube and implicated in LDA in thisstructure (Holtz et al., 2013). To test whether Ptch1 and Ptch2 areregulated by Shh in the limb, we cultured wildtype primary limbfibroblasts, harvested from the forelimbs of E12.5 wildtype embryos,with recombinant Shh protein at two different concentrations. Wedemonstrated that treatment with Shh results in a 4300-foldincrease in the level of Ptch2 transcripts (Fig. 3G), and 40-foldand 4200-fold increases in the levels of Ptch1 and Gli1 transcripts,respectively (Figs. 3H and I) indicating that Ptch2, like Ptch1, is highlyupregulated in the presence of Shh ligand. Importantly, we did not

observe increased signaling with higher concentration of Shh (1 μg/mL) suggesting that the lower concentration (0.1 μg/mL) is sufficientto achieve saturated signaling in wildtype limb fibroblasts.

AER-specific deletion of Ptch1 in the Ptch2 null background results insoft-tissue syndactyly

To test whether Ptch1 and Ptch2 share functional overlap in theShh-responsive tissues of the limb, we generated an AER andectoderm-specific conditional knockout of Ptch1, using Msx2-Crerecombinase, and examined limb patterning in a Ptch2 wildtype ornull background (Sun et al., 2000). Work by Bouldin et al. (2010)demonstrated that Shh signals to the ectoderm as evident fromPtch1-lacZ expression in the posterior AER. Consistent with thesefindings, we observe high levels of expression of Ptch1-lacZ andPtch2-lacZ in the posterior AER (Fig. 3A'–D'). To examine the role ofPtch1 and Ptch2 in the AER, we conditionally inactivated a Ptch1flox

(Ptch1f) allele, with Msx2-Cre, in a Ptch1-heterozygous background(either Ptch1� or Ptch1del alleles) (Ellis et al., 2003; Goodrich et al.,1997). This allowed us to examine the role of Ptch1 in the AER andthe ventral ectoderm, while tracing pathway activity and expres-sion of Ptch1 or Ptch2 using the lacZ reporter. Analysis of Msx2-Cre;Ptch1f/� limbs at E11.5 and E12.5 demonstrated that loss of Ptch1in the AER resulted in constitutive pathway activation with theexpression of Ptch1-lacZ throughout this structure, including in theanterior (Fig. S5A and B'). In Msx2-Cre;Ptch1f/d;Ptch2lacZ/þ mice, weshowed that there is a similar ectopic expression of Ptch2-lacZ inthe anterior AER, when Ptch1 is absent (Fig. S5C–D'). Together,these data support the hypothesis that Shh signaling is active inthe AER and pathway activation, through loss of Ptch1, results inelevated activation of Ptch1 and Ptch2 promoters.

Upregulation of both Ptch1 and Ptch2 in the mutant AER sug-gested that the two proteins may co-operate in regulating Shh

Fig. 2. Ptch2 regulates Smo localization and Shh signal transduction in Ptch1� /� MEFs. Smo (red) localization at primary cilia in Ptch1� /� MEFs (A) and upon over-expressionPtch1HA-IRESGFP (B) or Ptch2myc-IRESGFP (C). Primary cilia are marked by α-acetylated tubulin (green). Merged image (A'–C'). Percentage of cells with Smo localization at theprimary cilium was quantified (D). Retroviral over-expression of Ptch1HA-IRESGFP or Ptch2myc-IRESGFP also suppressed expression of Shh target genes Gli1 (E) and Hip(F) quantified by qPCR. (G–K) Smo localization (red) in Ptch1� /� MEFs (G) over-expressing Ptch1HA-IRESGFP (H-I) or Ptch2myc-IRESGFP (J and K) and treated with Shh proteinor carrier (BSA). (G'–K') Merged image highlights localization to α-acetylated tubulin positive (green) primary. Boxes highlight location of individual cilia. (L) Percentage ofcells with ciliary localization of Smo across treatments. (M) Treatment with Shh promotes activation of Gli1 expression in Ptch1HA-IRESGFP or Ptch2myc-IRESGFP expressingPtch1-/- MEFs. *Statistically significant difference, po0.05, Student's t-test, **Statistically significant difference, po0.01.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202194

signaling in the ectoderm. Previously, Bouldin et al. (2010) proposeda model for ectodermal–mesenchymal cross talk as a means toregulate the number and activity of Shh-expressing cells and theAER. According to this model, high Shh signaling in the AER isexpected to reduce Shh signaling in the posterior of the limb throughthe Shh/FGF/Gre loop. Although we did not observe a change in theexpression of AER marker Fgf8 at E11.5 (Fig. 4A–C), we detected areduction in the size of Shh positive domain (Fig. 4D and F) as well asan anterior shift in the expression of Gre in Msx2-Cre;Ptch1f/d;Ptch2lacZ/lacZ double mutants (Fig. 4G–I).

To determine how changes to Shh/FGF/Gre signaling affect limbdevelopment, we examined the phenotype of Msx2-Cre;Ptch1f/�

and Msx2-Cre;Ptch1f/d;Ptch2lacZ/lacZ mutants. Both single and dou-ble mutant embryos died at birth due to ectopic expression ofMsx2-Cre, and Shh pathway activation, in the head. This led to afailure to form parts of the parietal and frontal bones of the skulland the overlying ectoderm as well as defects in eye development(Fig. 5A–C). Analysis of skeletal patterning of Msx2-Cre;Ptch1f/-

and Msx2-Cre;Ptch1f/d;Ptch2lacZ/lacZ mutants at E18.5 revealed noanterior–posterior digit patterning defects in the forelimb or thehindlimb, suggesting that Ptch1 and Ptch2 expression in the AER isdispensable for AP pattern formation (Fig. 5D–I). However, thelimbs of double mutants were smaller than wildtype and exhibiteddelayed ossification at E18.5 (Fig. 5D–I) consistent with a role for

Shh/FGF signaling in limb outgrowth (Sun et al., 2002; Verheydenand Sun, 2008), although we cannot discount the possibility thatthis is secondary to growth retardation in the double mutants. Toassess whether the digit pattern was affected prior to ossification,we examined expression of Col2a1 mRNA as a marker of conden-sing mesenchyme in E12.5 limbs by RNA in situ hybridization. Wefound that at E12.5 the size and shape of limbs is grossly normal inthe single mutant (Fig. 5D'–G'), but double mutant limbs exhibit aslight delay Col2a1 expression, particularly in the hindlimb whereMsx2-Cre is expressed earlier (Fig. 5H' and I').

While the single and double mutant limbs did not exhibit severeskeletal defects, both genotypes developed soft tissue syndactyly(digit fusion/webbing), leading to a cupped or clasped appearance ofthe hands and feet (Fig. 5J–L). The phenotype was fully penetrant andboth Msx2-Cre;Ptch1f/- and Msx2-Cre;Ptch1f/d;Ptch2lacZ/lacZ mutantswere affected to the same degree. Analysis of the limbs betweenE12.75 and E14.5 revealed ectopic expression of Ptch2-lacZ on theventral side of mutant limbs indicating constitutive pathway activa-tion due to expression of Msx2-Cre (Fig. 5M and M'). Elevatedpathway activity did not affect the induction of Msx2 gene in theinterdigital mesenchyme, but its expression persisted longer in themutants compared to the wildtype (Fig. 5N and N') (Lallemandet al., 2009). Additionally, interdigital cell death was reduced anddelayed in the double mutant compared to wildtype (Fig. 5O and O').

Fig. 3. Ptch2 is upregulated in Shh-responding cells of the limb. (A–F) LacZ staining reveals Ptch1-lacZ and Ptch2-lacZ expression pattern in forelimb buds of reporter miceacross development. (G–I) Primary limb mesenchyme cells from E12.5 wildtype embryos were cultured in the presence of low (0.1 mg/mL) or high (1 mg/mL) concentration ofrecombinant Shh protein, or equivalent volume of BSA (carrier) for 24 h. Treatment with either concentration of Shh resulted in a 300-fold increase in Ptch2 mRNAexpression assessed by qPCR (G), and similar increases in level of Ptch1 (H) and Gli1 mRNA (I). **Statistically significant difference, po0.01 (Student's t-test).

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202 195

This suggested that constitutive pathway activation in the ventralectoderm may interfere with interdigital cell death resulting in softtissue digit fusions. Notably, there were no differences in the severityof the phenotype between the single and the double mutants,indicating that Ptch2 does not compensate for the loss of Ptch1 inthe ectoderm and that the phenotype may be driven solely by ectopicpathway activation due to loss of Ptch1 in the ventral ectoderm.Collectively, these data suggest that Ptch1 and Ptch2 expression in theAER is not required for normal AP patterning but promotes regula-tion of the Shh/FGF signaling loop and limb outgrowth.

Ptch1 and Ptch2 cooperate to regulate Smo in limb mesenchymein vitro and in vivo.

To determine whether Ptch1 and Ptch2 cooperate in patterningof the limb mesenchyme, we examined how loss of Ptch2 affectspathway activity in Ptch1-null fibroblasts derived from E12.5mutant and control limbs. Wild-type and Ptch2-/- limb fibroblasts

exhibit very low levels of ciliary localization of Smo (o2% of allciliated cells) (Fig. 6A and B', quantified in Fig. 6E). In contrast,Ptch1-null fibroblasts, derived from Prx1-Cre;Ptch1f/� mutantlimbs, exhibit constitutive pathway activation with ciliary localiza-tion of Smo in about 50% of the cells (Fig. 6C and C', quantified inFig. 6E). Notably, loss of Ptch2 in the Ptch1-null background (Prx1-Cre;Ptch1f/�;Ptch2� /�) resulted in a 20% increase in the number ofcells with Smoþ cilia (Fig. 6D and D', quantified in Fig. 6E). Theseresults indicate that, in the absence of Ptch1, Ptch2 regulates Smolocalization in the primary cilia. Consistent with this, both Prx1-Cre;Ptch1f/� and Prx1-Cre;Ptch1f/�;Ptch2� /� cells exhibit highlevels of pathway activation indicated by high levels of Gli1 andlow levels of Gli3R protein (Fig. 5F, lane 3–4). Notably, doublemutant cells exhibit a near complete absence of both full-lengthand truncated repressor isoforms of Gli3, the main transcriptionalrepressor of Shh signaling. These findings indicate a higher level ofpathway activation upon loss of both Ptch homologs and suggestthat Ptch2 cooperates with Ptch1 in negative regulation of Smo.

Fig. 4. Deletion of Ptch1 and Ptch2 in the AER perturbs the Shh-Fgf signaling loop. Whole-mount RNA in situ hybridization analysis ofMsx2-Cre;Ptch1f/� mutant limbs in a wildtypeor Ptch2 mutant background reveals no changes in the expression of Fgf8 (A-C). Decreased expression of Shh is observed upon loss of Ptch1 and Ptch2 (panel F in D–F). Ectopicexpression of Gremlin is observed in the anterior of the limb upon loss of Ptch1 (G–H) and this is further exacerbated upon loss of Ptch2 (I). All images show hindlimbs at E11.5.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202196

Ptch1 and Ptch2 play overlapping roles in digit outgrowthand specification.

To test whether the functional overlap between Ptch1 and Ptch2is biologically significant, we compared the phenotype of Prx1-Cre;Ptch1f/� and Prx1-Cre;Ptch1f/�;Ptch2� /� mutant forelimbs. We andothers have previously demonstrated that loss of Ptch1 results insevere oligodactyly and outgrowth defects in the embryonic limb(Fig. 7A and B) (Butterfield et al., 2009; Zhulyn et al., 2014).Notably, loss of Ptch2 in this background exacerbates the out-growth defect and completely abolishes digit formation in a dose-dependent manner (Fig. 7C and Fig. S6A–D'). Recently, we foundthat the outgrowth defects in the Prx1-Cre;Ptch1f/� forelimb aredue to precocious activation of Shh signaling which interferes withthe establishment of signaling centers and anterior progenitorspecification (Zhulyn et al., 2014). Here, we show that the Prx1-Cre;Ptch1f/�;Ptch2� /� mutant phenotype is also driven by thesame mechanism with constitutive pathway activation throughout

the limb exemplified by increased expression of Gli1 (Fig. 7D–F,data not shown). As in the Prx1-Cre;Ptch1f/� mutants, pathwayactivation is associated with failure to establish normal Shh andFgf8 signaling which drives limb outgrowth (Fig. 7G–L) (Butterfieldet al., 2009; Scherz et al., 2004; Zhulyn et al., 2014). Constitutivepathway activation also promotes ectopic induction of ‘posterior'markers, such as Hand2, and inhibits specification of anterioridentity, marked by Alx4, resulting in the re-specification of theentire mutant limb to a posterior fate (Fig. 7M–R).

To quantify how loss of Ptch2 affects pathway activity in vivo, wecompared the expression of Gli1 and Hip in E10.5 forelimbs fromPrx1-Cre;Ptch1f/� and Prx1-Cre;Ptch1f/-;Ptch2� /� mice and demon-strated a �2–3 fold further increase in pathway activity upon lossof Ptch2 (Fig. 6S and T) and a consistent increase in Gli1 protein anddecrease in the level of Gli3R (Fig. 6U). These findings illustrate that,in the absence of Ptch1, Ptch2 is a potent negative regulator of Shhsignaling, which plays a compensatory role in inhibiting Smo, and isinvolved in limb outgrowth and digit specification.

Fig. 5. Ptch1 and 2 activity in the AER is dispensable for normal skeletal pattern.Msx2-Crewas used to delete Ptch1 in the AER and ventral ectoderm in wildtype or Ptch2 nullbackground. (A–I) Alcian blue and Alizarin red staining of embryonic (E18.5) skeletons. (A–C) the heads of Ptch1 and Ptch1;Ptch2 mutants are small and exhibit a failure ofossification in the parietal and frontal bones due to ectopic expression of Msx2-Cre in the head and face. (D–H) The forelimbs and (E–I) hindlimbs of single and doublemutants are smaller than wildtype and exhibit delayed ossification, however anterior–posterior patterning of long bones and digits appears normal in mutants. RNA in situhybridization for Col2a1 shows no gross changes in the pattern of developing digits in (D'–I') the forelimbs (FL) and hindlimbs (HL) of single and double mutants at E12.5,however double mutants exhibit developmental delay. (J–L) Soft tissue syndactyly results in a clasped appearance of the forelimbs and hindlimbs in single and doublemutants at E18.5. Analysis of wildtype (M–O) and Ptch1;Ptch2 double mutant limbs (M'–O') demonstrates ectopic activation of Ptch2-lacZ in the anterior AER (red arrow) andventral ectoderm of the mutant limb (M'). Ectopic pathway activation is associated with normal expression of Msx2 (N') but delayed apoptosis of interdigital tissues (O')(white arrow).

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202 197

Ptch1 and Ptch2 play overlapping and complementary roles in Smoregulation

In this work, we demonstrate that Ptch2 is a functional receptorfor Shh and plays a critical role in the LIA of Smo. This functioncomplements its role in LDA and gradient regulation (Holtz et al.,2013). However, while Ptch2 exhibits similar regulatory propertiesand expression patterns as Ptch1, it is expressed at lower levelsthan Ptch1 throughout development, and is somewhat less effec-tive than Ptch1 at suppressing Smo or responding to Shh(Carpenter et al., 1998; Holtz et al., 2013; Motoyama et al.,1998a; Rahnama et al., 2004). Consistent with these differences,phylogenetic analysis indicates that Ptch1 and Ptch2 cluster intodistinct subgroups (Motoyama et al., 1998b). Their key differencelies in the shorter C-terminal tail of Ptch2, which lacks a conservedPPXY motif that promotes protein turnover, conferring a higherPtch2 protein stability (Carpenter et al., 1998; Kawamura et al.,2008; Lu et al., 2006). Together with our findings, these biochem-ical differences support the idea that Ptch1 is the primary Shhreceptor whereas Ptch2 plays a minor compensatory role in signaltransduction. In contrast, high level induction of Ptch2 in responseto Shh signaling as well as its lower turnover rate suggest thatPtch2 might play a more prominent role in LDA (Carpenter et al.,1998; Holtz et al., 2013; Kawamura et al., 2008). Though Ptch1 andPtch2 are highly expressed in response to pathway activation, thegreater stability of Ptch2 (Kawamura et al., 2008) coupled with itsapparently lower affinity for the ligand (Kd 1.8 nM for Ptch2 and1.0 nM for Ptch1) (Carpenter et al., 1998), and its capacity totransduce Shh suggest that Ptch2 might contribute to the sensingand transducing the Shh gradient over time and space. In addition,Ptch1 and Ptch2 belong to the RND (resistance nodulation divi-sion) family of proteins. Some members of this family, includingPtch1, are thought to form trimeric complexes (Liu et al., 2006,2009; Tseng et al., 1999). A compelling hypothesis is that upregu-lation of Ptch2, in response to Shh signaling, might alter thecomposition of Ptch complexes at the cell surface, which couldregulate the spread of Shh and/or alter the cellular responsivenessto the Shh gradient.

Until recently, Ptch2 was an unappreciated component of theShh signaling cascade. Work from our laboratory and others hashighlighted its importance in LDA in the neural tube (Holtz et al.,2013) and pathway regulation in the skin (Adolphe et al., 2014;Nieuwenhuis et al., 2006), however little was known about itsbiochemical properties. This work attempts to examine theoverlapping roles of Ptch1 and Ptch2 in vivo and to highlighttheir common function in regulating Smo. While this work was inrevision, a critically important analysis of the role of Ptch2 in Shh-responsiveness of Ptch1-/- MEFs and neuralized embryoid bodieswas published (Alfaro et al., 2014). This study, together with ourwork, reveals that Ptch2 is an important regulator of Shhsignaling.

Materials and methods

Generation of Ptch1HA and Ptch2myc expressing cell lines

Ptch1HA and Ptch2myc cDNA was subcloned into the pMIGmammalian retroviral expression vector and used to infect Ptch1� /

� MEFs (Taipale et al., 2000). pMIG contains an IRESGFP which wasused for fluorescent sorting of infected cells. Ptch1HA or Ptch2mycexpression in GFPþ cell cultures was assessed by Western blotanalysis. Ptch1 was detected using a goat anti-Ptch1 polyclonalantibody (1:500 dilution, G-19, sc-6149, Santa Cruz), Ptch2 wasdetected using a rabbit anti-Ptch2 antibody (1:1000 dilution, L849,Cell Signaling), HA was detected using a mouse monoclonal anti-body (1:2000 dilution, F7, Santa Cruz), and c-myc was detectedusing a rabbit polyclonal antibody (1:1000 dilution, A-14, sc-789,Santa Cruz).

Mouse lines and embryo analysis

Analysis was performed using the following strains Ptch2-

(Ptch2tm1) (Nieuwenhuis et al., 2006), Ptch2lacZ (Adolphe et al.,2014), Ptch1flox and Ptch1del (Ellis et al., 2003), Ptch1lacZ (Goodrichet al., 1997), Prx1-Cre (Logan et al., 2002), Msx2-Cre (Sun et al.,

Fig. 6. Ptch2 regulates Smo activity in vivo: (A–D) Merged images show Smo (green) localization to primary cilia marked by α-acetylated tubulin (red) in limb fibroblastsderived from forelimbs of E12.5 wildtype and mutant embryos and examined at confluency. (A'–D') Location of cilia is highlighted by α-acetylated tubulin (red). Nuclei arestained with DAPI (blue). (E) Ptch1;Ptch2 double mutant cells exhibit significantly higher percentage of Smoþ primary cilia compared to Ptch1 single mutants. (F) Gli1 proteinlevel is increased, Gli2 is unchanged and Gli3 is not detected, in Ptch1;Ptch2 double mutant cells. Lane 1 – wildtype; lane 2 – Ptch2 mutant; lane 3 – Ptch1mutant; and lane 4– double mutant limb fibroblast lysates.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202198

2002). Mouse lines were maintained on an outbred CD1 back-ground and all animal work was performed in compliance withprotocols and regulations established by the Hospital for SickChildren Animal Care Committee. For embryonic analysis and cellharvest we set-up timed matings, with the day of the vaginal plugscored as E0, and harvested embryos into PBS at the specifiedstages. Embryos were fixed in 4% paraformaldehyde (PFA) for RNA

in situ hybridization, in 80% ethanol for skeletal staining, or inX-gal fixative solution (2.7% formaldehyde (252549-500ML,Sigma), 0.02% Nonidet-P40 (BDH), 0.1 M phosphate buffered saline(PBS)) for whole-mount lacZ staining. Fresh limb buds weredissected on ice and snap frozen in liquid nitrogen, for subsequentanalysis by Western Blot or qPCR, or dissociated for primary cellculture.

Fig. 7. Ptch2 is required for digit patterning and outgrowth in the absence of Ptch1. (A–C) Col2a1 in situ hybridization indicates progressive truncation of the limb and digitloss upon deletion of Ptch1 and 2. (D–F) Ectopic expression of Gli1 throughout the limb mesenchyme observed upon loss of Ptch1 or Ptch1 and Ptch2. * Prx1-Cre;Ptch1f/�;Ptch2þ /� littermate shown for comparison (E). (G–L) Abnormal expression of Shh at the periphery of the limb, and loss of Fgf8 in the AER, observed in the double mutant andPtch1-mutant control. (M–R) Ectopic expression of Hand2 and loss of Alx4 indicate abnormal AP patterning in the single and double mutant. Analysis of E10.5 forelimb mRNAindicates that loss of Ptch2 results in increased expression of Gli1 (S) and Hip (T) in the absence of Ptch1. (U) Western Blot analysis of forelimb lysates indicates elevated Gli1and loss of Gli3R in Ptch1 and Ptch1;Ptch2 at E11.5.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202 199

Primary limb fibroblast culture

Forelimb buds of mutant (Prx1-Cre;Ptch1f/� , Prx1-Cre;Ptch1f/�;Ptch2� /� and Ptch2� /�) and control embryos, at E12.5, weredissected out and minced into small pieces in 1� PBS. The tissuewas incubated in 1� Trypsin–EDTA (TE) (15400-054, Gibco) for15 min at 37 1C, pelleted in 15%FBS–DMEM (FBS F2442-500ML,Sigma; DMEM 11965-092, Gibco; penicillin/streptomycin 15140-122, Gibco) and plated on 96-well plates. After expansion, 2.1�105

cells were seeded on gelatin-coated cover slips and allowed togrown to confluence in 15%FBS–DMEM before immunostainingwas performed as described.

Immunostaining and Smo localization

Ptch1� /� mouse embryonic fibroblasts (MEFs), over-expressingPtch1HA or Ptch2myc, or primary limb fibroblasts, were seeded ongelatin-coated cover slips and grown to confluence in 10%FBS–DMEM,or 15%FBS–DMEM media with penicillin/streptomycin. Cells weremoved to “starvation” media (0.5%FBS–DMEM, with penicillin/strep-tomycin) during the final 48 h of growth and treated with ShhNprotein (0.1 μg/mL or 1 μg/mL) (rmShh C2511-N R&D 464-5H, Cedar-lane), SAG (0.1 μm or 1 μM, Santa Cruz: sc-202814), or carrier in 0.5%FBS–DMEM in the final 24 h. Immunostaining was performed follow-ing F. Charron, with cells fixed on coverslips in 4%PFA at 4 1C, blockedin 1�PBS with 10% goat serum (CL1200, Cedarlane) and 0.1% TritonX-100 (X100-1L, Sigma) for 1 h. Samples were incubated with primaryantibodies against Smo (1:100, LS-A2668, MBL International Corpora-tion) and α-acetylated tubulin (1:1000, T6793, Sigma) overnight at4 1C, washed in 1� PBS with 1% goat serum and 0.1%TritonX-100, andincubated with TRITC, FITC, or Alexa488 (all 1:1000 dilution; 111-025-144 and 115-096-003 Jackson Laboratories; A11034, Invitrogen) sec-ondary antibodies for 1 h at room temperature. Samples were washedand mounted in media containing DAPI (H-1200, Vector) and imagedusing a Zeiss Axiovert 200M inverted fluorescence microscope. Toquantify changes in Smo localization, we performed at least threeindependent replicates of each experiment. 10–20 fields of view wererandomly selected and imaged for each slide and Smo localizationwasassessed. Statistical significance was determined using Student's t-test.

Phenotype analysis: whole-mount lacZ staining, in situ hybridizationand skeletal staining

Samples were fixed in X-gal fixative for 15 min (E9.5) toovernight (4E12.5), washed in X-gal buffer (1�PBS, 0.002 MMgCl2, and 0.02% Nonidet-P40) and stained overnight at 37 1C instaining solution containing 2 mM MgCl2, 0.02% Nonidet-P40,5 mM K4Fe(CN)6 �3H2O (P-8131, Sigma), 5 mM K3Fe(CN)6(P-9387, Sigma), and 50 mg/mL X-gal in dimethylformamide. Thesamples were washed in 1�PBS and cleared and imaged inmethanol. For RNA in situ hybridization, samples were fixedovernight in 4%PFA and processed as described (Nieuwenhuiset al., 2006). Hybridization was performed using the followingdioxigenin-labelled RNA probes: Col2a1 (M. Metsaranta), Shh(A. McMahon), Fgf8 (A. Tanaka), Gli1 (A. Joyner), Alx4 (H. Masuya),Hand2 (M. Scott). Samples were cleared in methanol and imagedusing a Leica Fluorescence Stereoscope. For skeletal staining,samples were collected at E18.5 and P2 and processed followingstandard protocols as previously described (Nieuwenhuis et al.,2006). Limbs were imaged in 50% glycerol:50% ethanol.

Cell sorting and in vivo lacZ assay

Homozygous Ptch2lacZ/lacZ animals were crossed to wildtype(CD1) to generate litters of heterozygous Ptch2lacZ/þ embryos.Litters were dissected at E10.5 and limbs were dissociated by

mechanical separation followed by treatment with 1�TE (20 minat 37 1C). Following resuspension in 15%FBS–DMEM, cells werepelleted and in vivo lacZ assay was performed following manu-facturer's instructions using the Marker Gene in vivo lacZß-Galactosidase Intracellular Detection Kit (Marker Gene Technol-ogies, Inc. M0259). Samples were incubated with FDG reagent for25 min then pelleted, re-suspended in sorting buffer (1%FBS, 1 mMEDTA, 1�PBS) and filtered. Propidium iodide (PI) (P3566, Invitro-gen) was added to a final dilution of 1:1000. Samples were sortedinto lacZ positive and negative fractions using a BD FACS Aria II cellsorter (BD Biosciences) and collected into 15%FBS–DMEM. Aftercollection, samples were pelleted and resuspended in 500 μL ofTrizol. RNA extraction was performed following standard methods(Zhulyn et al., 2014).

Western blot and qPCR

Western Blot analysis was carried out using previouslyreported methods (Zhulyn et al., 2014). For detection of Ptch1and Ptch2 protein preparations were incubated with ß-mercaptoethanol at 37 ºC for 15 min, for detection of Gli1, Gli2and Gli3 incubation was performed at 95 ºC for 5 min. Protein wasdetected using a rabbit polyclonal antibody against Gli1 (1:1000dilution, V812, Cell Signaling), rabbit polyclonal antibody againstGli2 (1:1000 dilution, Vijitha Puviindran, Hui Lab) (Zhulyn et al.,2014), goat polyclonal antibody against Gli3 (1:1000 dilution,AF3690, R&D), and mouse monoclonal antibody against Actin(1:10,000 dilution, CPO1 Ab-1, Calbiochem). RNA was extractedwith Trizol reagent (15596018, Ambion) as described previously(Li et al., 2012). Each experiment utilized at least three indepen-dent biological replicates and was repeated at least two times.qPCR was carried out using TaqMan Universal PCR mastermix (4324018, Applied Biosystems) and TaqMan Gene Expres-sion Assay probes for mouse Gli1 (Mm00494646_m1), Hip(Mm00469580_m1), Ptch1 and Ptch2. Gapdh (Mm03302249_g1)was used as the endogenous control for normalization. Geneexpression was assessed using the 2ΔΔCt method and statisticalsignificance was confirmed using Student's t-test as describedpreviously (Zhulyn et al., 2014).

Acknowlegements

We thank Dr. Brandon Wainwright for his kind gift of Ptch1flox

mice. We thank The Imaging Facility at the Hospital for SickChildren and the SickKids – UHN Flow Cytometry Facility for useof equipment and training. This work was supported by a CanadianInstitutes of Health Research Grant (MOP 86665) to C.C.H., andUniversity of Toronto Fellowship and Doctoral Completion Grant toO.Z. The project was conceived by E.N., O.Z. and C.C.H. All in vitroand in vivo experiments were conceived, designed and carried outby O.Z. Retroviral infection was carried out by Y.C.L. and S.A.Ptch2lacZ mice were generated by E.N. The manuscript was writtenand edited by O.Z. and C.C.H. The authors declare no competinginterests.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.ydbio.2014.10.023.

References

Adolphe, C., Nieuwenhuis, E., Villani, R., Li, Z.J., Kaur, P., Hui, C.-C., Wainwright, B.J.,2014. Patched 1 and patched 2 redundancy has a key role in regulatingepidermal differentiation. J. Invest. Dermatol., 1–10.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202200

Alfaro, A.C., Roberts, B., Kwong, L., Bijlsma, M.F., Roelink, H., 2014. Ptch2 mediatesthe Shh response in Ptch1 -/- cells. Development, 1–9.

Allen, B.L., Tenzen, T., McMahon, A.P., 2007. The Hedgehog-binding proteins Gas1and Cdo cooperate to positively regulate Shh signaling during mouse develop-ment. Genes Dev. 21, 1244–1257.

Allen, B.L., Song, J.Y., Izzi, L., Althaus, I.W., Kang, J.-S., Charron, F., Krauss, R.S.,McMahon, A.P., 2011. Overlapping roles and collective requirement for thecoreceptors GAS1, CDO, and BOC in SHH pathway function. Dev. Cell 20,775–787.

Bae, G.-U., Domené, S., Roessler, E., Schachter, K., Kang, J.-S., Muenke, M., Krauss, R.S.,2011. Mutations in CDON, encoding a hedgehog receptor, result in holoprosen-cephaly and defective interactions with other hedgehog receptors. Am. J. Hum.Genet. 89, 231–240.

Bouldin, C.M., Gritli-Linde, A., Ahn, S., Harfe, B.D., 2010. Shh pathway activation ispresent and required within the vertebrate limb bud apical ectodermal ridgefor normal autopod patterning. Proc. Natl. Acad. Sci. USA 107, 5489–5494.

Briscoe, J., Chen, Y., Jessell, T.M., Struhl, G., 2001. A hedgehog-insensitive form ofpatched provides evidence for direct long-range morphogen activity of sonichedgehog in the neural tube. Mol. Cell 7, 1279–1291.

Butterfield, N.C., Metzis, V., McGlinn, E., Bruce, S.J., Wainwright, B.J., Wicking, C.,2009. Patched 1 is a crucial determinant of asymmetry and digit number in thevertebrate limb. Development 136, 3515–3524.

Carpenter, D., Stone, D.M., Brush, J., Ryan, A., Armanini, M., Frantz, G., Rosenthal, A.,de Sauvage, F.J., 1998. Characterization of two patched receptors for thevertebrate hedgehog protein family. Proc. Natl. Acad. Sci. USA 95, 13630–13634.

Chen, Y., Struhl, G., 1996. Dual roles for patched in sequestering and transducingHedgehog. Cell 87, 553–563.

Chen, J.K., Taipale, J., Young, K.E., Maiti, T., Beachy, P.A., 2002. Small moleculemodulation of Smoothened activity. Proc. Natl. Acad. Sci. USA 99, 14071–14076.

Chiang, C., Litingtung, Y., Harris, M.P., Simandl, B.K., Li, Y., Beachy, P.A., Fallon, J.F.,2001. Manifestation of the limb prepattern: limb development in the absenceof sonic hedgehog function. Dev. Biol. 236, 421–435.

Chuang, P., Mcmahon, A.P., 1999. Vertebrate Hedgehog signalling modulated byinduction of a Hedgehog-binding protein. Nature 397, 617–621.

Corbit, K.C., Aanstad, P., Singla, V., Norman, A.R., Stainier, D.Y.R., Reiter, J.F., 2005.Vertebrate Smoothened functions at the primary cilium. Nature 437, 1018–1021.

Dorn, K.V., Hughes, C.E., Rohatgi, R., 2012. A smoothened-Evc2 complex transducesthe Hedgehog signal at primary cilia. Dev. Cell 23, 823–835.

Ellis, T., Smyth, I., Riley, E., Graham, S., Elliot, K., Narang, M., Kay, G.F., Wicking, C.,Wainwright, B., 2003. Patched 1 conditional null allele in mice. Genesis 36,158–161.

Goodrich, L.V., Johnson, R.L., Milenkovic, L., McMahon, J.A., Scott, M.P., 1996.Conservation of the hedgehog/patched signaling pathway from flies to mice:induction of a mouse patched gene by Hedgehog. Genes Dev. 10, 301–312.

Goodrich, L.V., Milenkovic, L., Higgins, K.M., Scott, M.P., 1997. Altered neural cellfates and medulloblastoma in mouse patched mutants. Science 277, 1109–1113.

Holtz, A.M., Peterson, K.a., Nishi, Y., Morin, S., Song, J.Y., Charron, F., McMahon, A.P.,Allen, B.L., 2013. Essential role for ligand-dependent feedback antagonism ofvertebrate hedgehog signaling by PTCH1, PTCH2 and HHIP1 during neuralpatterning. Development 140, 3423–3434.

Hooper, J.E., Scott, M.P., 1989. The Drosophila patched gene encodes a putativemembrane protein required for segmental patterning. Cell 59, 751–765.

Hui, C.-C., Angers, S., 2011. Gli proteins in development and disease. Annu. Rev. CellDev. Biol. 27, 513–537.

Ingham, P.W., Nystedt, S., Nakano, Y., Brown, W., Stark, D., van den Heuvel, M.,Taylor, A.M., 2000. Patched represses the Hedgehog signalling pathway bypromoting modification of the Smoothened protein. Curr. Biol. 10, 1315–1318.

Izzi, L., Lévesque, M., Morin, S., Laniel, D., Wilkes, B.C., Mille, F., Krauss, R.S.,McMahon, A.P., Allen, B.L., Charron, F., 2011. Boc and Gas1 each form distinctShh receptor complexes with Ptch1 and are required for Shh-mediated cellproliferation. Dev. Cell 20, 788–801.

Jeong, J., McMahon, A.P., 2005. Growth and pattern of the mammalian neural tubeare governed by partially overlapping feedback activities of the hedgehogantagonists patched 1 and Hhip1. Development 132, 143–154.

Jiang, J., Hui, C.-C., 2008. Hedgehog signaling in development and cancer. Dev. Cell15, 801–812.

Kavran, J.M., Ward, M.D., Oladosu, O.O., Mulepati, S., Leahy, D.J., 2010. Allmammalian hedgehog proteins interact with CDO and BOC in a conservedmanner. J. Biol. Chem., 1–17.

Kawamura, S., Hervold, K., Ramirez-Weber, F.-A., Kornberg, T.B., 2008. Two patchedprotein subtypes and a conserved domain of group I proteins that regulatesturnover. J. Biol. Chem. 283, 30964–30969.

Koudijs, M.J., den Broeder, M.J., Groot, E., van Eeden, F.J., 2008. Genetic analysis ofthe two zebrafish patched homologues identifies novel roles for the hedgehogsignaling pathway. BMC Dev. Biol. 8, 15.

Kraus, P., Fraidenraich, D., Loomis, C.A., 2001. Some distal limb structures develop inmice lacking Sonic hedgehog signaling. Mech. Dev. 100, 45–58.

Lallemand, Y., Bensoussan, V., Cloment, C., Saint, Robert, B., 2009. Msx genes areimportant apoptosis effectors downstream of the Shh/Gli3 pathway in the limb.Dev. Biol. 331, 189–198.

Lee, C.S., Fan, C.M., 2001. Embryonic expression patterns of the mouse and chickGas1 genes. Mech. Dev. 101, 293–297.

Lee, C.S., Buttitta, L., Fan, C.M., 2001. Evidence that the WNT-inducible growtharrest-specific gene 1 encodes an antagonist of sonic hedgehog signaling in thesomite. Proc. Natl. Acad. Sci. USA 98, 11347–11352.

Lee, Y., Miller, H.L., Russell, H.R., Boyd, K., Curran, T., McKinnon, P.J., 2006. Patched2modulates tumorigenesis in patched1 heterozygous mice. Cancer Res. 66,6964–6971.

Li, Z.J., Nieuwenhuis, E., Nien, W., Zhang, X., Zhang, J., Puviindran, V., Wainwright, B.J.,Kim, P.C.W., Hui, C.-C., 2012. Kif7 regulates Gli2 through Sufu-dependentand -independent functions during skin development and tumorigenesis. Develop-ment 139, 4152–4161.

Liu, R., Lu, P., Chu, J.W.K., Sharom, F.J., 2009. Characterization of fluorescent sterolbinding to purified human NPC1. J. Biol. Chem. 284, 1840–1852.

Logan, M., Martin, J.F., Nagy, A., Lobe, C., Olson, E.N., Tabin, C.J., 2002. Expression ofCre Recombinase in the developing mouse limb bud driven by a Prxl enhancer.Genesis 33, 77–80.

Lu, X., Liu, S., Kornberg, T.B., 2006. The C-terminal tail of the Hedgehog receptorPatched regulates both localization and turnover. Genes Dev. 20, 2539–2551.

Marigo, V., Scott, M.P., Johnson, R.L., Goodrich, L.V., Tabin, C.J., 1996. Conservation inhedgehog signaling: induction of a chicken patched homolog by Sonic hedge-hog in the developing limb. Development 122, 1225–1233.

Martinelli, D.C., Fan, C.-M., 2007. Gas1 extends the range of Hedgehog action byfacilitating its signaling. Genes Dev. 21, 1231–1243.

Milenkovic, L., Goodrich, L.V., Higgins, K.M., Scott, M.P., 1999. Mouse patched1controls body size determination and limb patterning. Development 126,4431–4440.

Milenkovic, L., Scott, M.P., Rohatgi, R., 2009. Lateral transport of Smoothened fromthe plasma membrane to the membrane of the cilium. J. Cell Biol. 187, 365–374.

Motoyama, J., Heng, H., Crackower, M.A., Takabatake, T., Takeshima, K., Tsui, L.-C.,Hui, C.-C., 1998a. Overlapping and non-overlapping Ptch2 expression with Shhduring mouse embryogenesis. Mech. Dev. 78, 81–84.

Motoyama, J., Takabatake, T., Takeshima, K., Hui, C.-C., 1998b. Ptch2, a secondmouse Patched gene is co-expressed with Sonic hedgehog. Nat. Genet. 18,104–106.

Nieuwenhuis, E., Motoyama, J., Barnfield, P.C., Yoshikawa, Y., Zhang, X., Mo, R.,Crackower, M.A., Hui, C.-C., 2006. Mice with a targeted mutation of patched2are viable but develop alopecia and epidermal hyperplasia. Mol. Cell. Biol. 26,6609–6622.

Okada, A., Charron, F., Morin, S., Shin, D.S., Wong, K., Fabre, P.J., Tessier-Lavigne, M.,McConnell, S.K., 2006. Boc is a receptor for sonic hedgehog in the guidance ofcommissural axons. Nature 444, 369–373.

Pearse, R.V., Vogan, K.J., Tabin, C.J., 2001. Ptc1 and Ptc2 transcripts provide distinctreadouts of Hedgehog signaling activity during chick embryogenesis. Dev. Biol.239, 15–29.

Rahnama, F., Toftgård, R., Zaphiropoulos, P.G., 2004. Distinct roles of PTCH2 splicevariants in Hedgehog signalling. Biochem. J. 378, 325–334.

Rana, R., Carroll, C.E., Lee, H.-J., Bao, J., Marada, S., Grace, C.R.R., Guibao, C.D., Ogden, S.K.,Zheng, J.J., 2013. Structural insights into the role of the Smoothened cysteine-richdomain in Hedgehog signalling. Nat. Commun 4, 2965.

Rohatgi, R., Milenkovic, L., Scott, M.P., 2007. Patched1 regulates hedgehog signalingat the primary cilium. Science 317, 372–376.

Rohatgi, R., Milenkovic, L., Corcoran, R.B., Scott, M.P., 2009. Hedgehog signaltransduction by Smoothened : Pharmacologic evidence for a 2-step activationprocess. Proc. Natl. Acad. Sci. USA 106, 3196–3201.

Sanders, T.A., Llagostera, E., Barna, M., 2013. Specialized filopodia direct long-range transport of SHH during vertebrate tissue patterning. Nature 497,628–632.

Scherz, P.J., Harfe, B.D., McMahon, A.P., Tabin, C.J., 2004. The limb bud Shh-Fgffeedback loop is terminated by expansion of former ZPA cells. Science 305,396–399.

Scherz, P.J., McGlinn, E., Nissim, S., Tabin, C.J., 2007. Extended exposure to Sonichedgehog is required for patterning the posterior digits of the vertebrate limb.Dev. Biol. 308, 343–354.

Seppala, M., Depew, M.J., Martinelli, D.C., Fan, C., Sharpe, P.T., Cobourne, M.T., 2007.Gas1 is a modifier for holoprosencephaly and genetically interacts with sonichedgehog. J. Clin. Invest. 117, 1575–1584.

Sun, X., Lewandoski, M., Meyers, E.N., Liu, Y.H., Maxson, R.E., Martin, G.R., 2000.Conditional inactivation of Fgf4 reveals complexity of signalling during limbbud development. Nat. Genet. 25, 83–86.

Sun, X., Mariani, F.V., Martin, G.R., 2002. Functions of FGF signalling from the apicalectodermal ridge in limb development. Nature 418, 501–508.

Taipale, J., Chen, J.K., Cooper, M.K., Wang, B., Mann, R.K., Milenkovic, L., Scott, M.P.,Beachy, P.A., 2000. Effects of oncogenic mutations in Smoothened and Patchedcan be reversed by cyclopamine. Nature 406, 1005–1009.

Tenzen, T., Allen, B.L., Cole, F., Kang, J.-S., Krauss, R.S., McMahon, A.P., 2006. The cellsurface membrane proteins Cdo and Boc are components and targets of theHedgehog signaling pathway and feedback network in mice. Dev. Cell 10,647–656.

Tseng, T.T., Gratwick, K.S., Kollman, J., Park, D., Nies, D.H., Goffeau, A., Saier, M.H.,1999. The RND permease superfamily: an ancient, ubiquitous and diversefamily that includes human disease and development proteins. J. Mol. Micro-biol. Biotechnol. 1, 107–125.

Verheyden, J.M., Sun, X., 2008. An Fgf/Gremlin inhibitory feedback loop triggerstermination of limb bud outgrowth. Nature 454, 638–641.

Wang, Y., Zhou, Z., Walsh, C.T., McMahon, A.P., 2009. Selective translocation ofintracellular Smoothened to the primary cilium in response to Hedgehogpathway modulation. Proc. Natl. Acad. Sci. USA 106, 2623–2628.

Wilson, C.W., Chen, M.-H., Chuang, P.-T., 2009. Smoothened adopts multiple activeand inactive conformations capable of trafficking to the primary cilium. PLoSONE 4, e5182.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202 201

Zhang, W., Kang, J.-S., Cole, F., Yi, M.-J., Krauss, R.S., 2006. Cdo functions at multiplepoints in the Sonic Hedgehog pathway, and Cdo-deficient mice accuratelymodel human holoprosencephaly. Dev. Cell 10, 657–665.

Zhang, W., Hong, M., Bae, G., Kang, J.-S., Krauss, R.S., 2011. Boc modifies the holo-prosencephaly spectrum of Cdo mutant mice. Dis. Model. Mech. 4, 368–380.

Zhulyn, O., Li, D., Deimling, S., Vakili, N.A., Mo, R., Puviindran, V., Chen, M.-H.,Chuang, P.-T., Hopyan, S., Hui, C., 2014. A switch from low to high Shh activityregulates establishment of limb progenitors and signaling centers. Dev. Cell,1–9.

O. Zhulyn et al. / Developmental Biology 397 (2015) 191–202202