- Email: [email protected]
Functional domains and sub-cellular distribution of the Hedgehog transducing protein Smoothened in Drosophila
Mechanisms of Development 121 (2004) 507–518 www.elsevier.com/locate/modo
Functional domains and sub-cellular distribution of the Hedgehog transducing protein Smoothened in Drosophila Y. Nakano1,2, S. Nystedt1, A.A. Shivdasani1, H. Strutt1, C. Thomas1, P.W. Ingham* MRC Intercellular Signalling Group, Centre for Developmental Genetics, University School of Medicine and Biomedical Science, Firth Court, Western Bank, Sheffield S10 2TN, UK Received 17 September 2003; received in revised form 30 March 2004; accepted 15 April 2004 Available online 28 April 2004
Abstract The Hedgehog signalling pathway is deployed repeatedly during normal animal development and its inappropriate activity is associated with various tumours in human. The serpentine protein Smoothened (Smo) is essential for cells to respond to the Hedeghog (Hh) signal; oncogenic forms of Smo have been isolated from human basal cell carcinomas. Despite similarities with ligand binding G-protein coupled receptors, the molecular basis of Smo activity and its regulation remains unclear. In non-responding cells, Smo is suppressed by the activity of another multipass membrane spanning protein Ptc, which acts as the Hh receptor. In Drosophila, binding of Hh to Ptc has been shown to cause an accumulation of phosphorylated Smo protein and a concomitant stabilisation of the activated form of the Ci transcription factor. Here, we identify domains essential for Smo activity and investigate the sub-cellular distribution of the wild type protein in vivo. We find that deletion of the amino terminus and the juxtamembrane region of the carboxy terminus of the protein result in the loss of normal Smo activity. Using Green Fluorescent Protein (GFP) and horseradish peroxidase fusion proteins we show that Smo accumulates in the plasma membrane of cells in which Ptc activity is abrogated by Hh but is targeted to the degradative pathway in cells where Ptc is active. We further demonstrate that Smo accumulation is likely to be a cause, rather than a consequence, of Hh signal transduction. q 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Smoothened; Oncogene; Signalling; Patched; Hedgehog; Drosophila
1. Introduction Despite substantial knowledge of the plethora of processes in which Hedeghog (Hh) signals are deployed during animal development and the growing recognition of the role played by aberrant Hh signalling in various human tumours (see McMahon et al. (2003)), many gaps remain in our understanding of how the Hh signal is transduced. In particular, the mechanisms by which the Smoothened (Smo) protein—an obligate component of the pathway (Alcedo et al., 1996; van den Heuvel and Ingham, 1996; Chen et al., 2001; Varga et al., 2001; Zhang et al., 2001) and a target for oncogenic mutation (Xie et al., 1998)—is regulated and activates its intracellular targets are still largely obscure. Although it shares structural similarities with serpentine receptor proteins, particularly members of the Frizzled * Corresponding author. Tel.: þ44-114-222-2803; fax: þ 44-114-2222788. E-mail address: [email protected] (P.W. Ingham). 1 These authors made equal contributions and are listed in alphabetical order. 2 Present address: Department of Genetics, Hyogo College of Medicine, Mukogawa-cho, 1-1, Nishinomiya 663-8501, Japan. 0925-4773/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mod.2004.04.015
family of Wnt receptors (Quirk et al., 1997) genetic and biochemical data indicate that Smo is not directly regulated by Hh ligand. Instead, its activity is somehow controlled by the multipass transmembrane protein Patched (Ptc), which itself acts as the Hh receptor. In unstimulated cells Ptc represses Smo activity; binding of Hh to Ptc relieves this repression, allowing Smo to signal to a multimeric complex inside the cell that includes the transcription factor Ci. In the absence of Hh signalling, Ci is sequestered in a cytoplasmic complex that includes the kinesin related protein Cos-2 and the serine threonine kinase Fused. This sequestration inhibits nuclear import of the full-length 155 kDa protein and promotes its cleavage to a 75 kDa form that can enter the nucleus and repress transcription. Activation of Smo inhibits Ci cleavage and activates the full-length protein, resulting in the transcription of Hh-responsive target genes (reviewed in Ingham and McMahon (2001)). The way in which the activity of Smo is transmitted to the Ci-Cos-2-Fu complex is obscure. Whilst Smo has a predicted heptahelical domain (Alcedo et al., 1996; van den Heuvel and Ingham, 1996), the signature motif of G-protein coupled receptors (GPCRs), there is no reported primary sequence similarity with any bona fide GPCR.
508
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Activity of the vertebrate Smo protein can be modulated by the alkaloid cyclopamine and other small molecule antagonists and agonists that have been shown to bind to the heptahelical domain (Chen et al., 2002a,b). Exactly how these small molecules influence Smo activity is unclear, but one suggestion is that they may alter the conformation of the protein, respectively, towards a more active or inactive state (Chen et al., 2002b). In this view, the normal regulation of Smo by Ptc might involve the transport of a cellular small molecule agonist/antagonist that similarly binds to Smo, altering its conformation. In line with this scenario, Ptc has been shown to share essential structural motifs with a family of bacterial permeases. Mutations in these motifs lead to a loss of its Smo suppressing activity (Taipale et al., 2002). Such permease activity would also be consistent with a role for Ptc in redistribution of Smo to a membrane compartment where it is susceptible to such an antagonist ligand or inaccessible to an agonist. In this view, the accumulation of Smo in the appropriate cellular location would be the key determinant of Smo activation.
Here, we present the results of in vivo analyses in Drosophila designed to investigate the functional domains and sub-cellular distribution of the Smo protein. Through both in vivo and in vitro mutagenesis, we have identified regions of the protein critical for its function. Using fusion proteins we have studied the changes in sub-cellular distribution of Smo in response to Hh and Ptc activity in vivo. Our results reveal critical requirements for both the amino and carboxy terminal domains for normal Smo activity and suggest that Ptc regulates Smo activity at least in part by targeting it to the lysosome.
2. Results 2.1. The sub-cellular distribution of Smo is regulated by Hh and Ptc activity Previous studies have shown that Smo protein accumulates preferentially in Hh responding cells in the Drosophila
Fig. 1. Regulation of Smo distribution in the wing imaginal discs and larval ectoderm. A GFP-tagged form of Smo expressed in wing imaginal discs under the control of the ap-GAL4 driver, accumulates in cells of the posterior compartment and just anterior to the boundary (right) but becomes degraded in more anterior cells (left) that do not receive the Hh signal (a, a0 ). By contrast, the distribution of GFP < SmoA479Y, a putative gain of function, tagged Smo mutant, does not appear to be regulated; the protein accumulating equally in anterior and posterior compartments (b, b0 ); the expression of dpp-lacZ (red) marks the anterior/posterior compartment boundary and ectopic Hh signal transduction in the anterior compartment; in both (a) and (b) dorsal is up. (c–e) Fluorescently tagged Smo molecules are post-transcriptionally regulated in larval ectoderm cells when expressed ubiquitously with act-GAL4 (top panel). High magnification projected images of several confocal sections (d, e) indicate that the tagged Smo accumulates predominantly at the plasma membrane where it appears to aggregate into clusters. (f–h) Simultaneous misexpression of both YFP < Smo and CFP < Ptc, results in the loss of Smo from the plasma membrane into intracellular vesicles. Occasionally, co-localisation of CFP < Ptc and YFP < Smo can be observed (arrow).
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
embryo and imaginal discs; moreover, in clone 8 cells the proportion of Smo present at the cell surface increases on exposure to Hh protein (Denef et al., 2000). To analyse the modulation of Smo distribution by Hh in vivo we generated a transgene encoding a GFP tagged form of the protein and expressed this in transgenic animals using the GAL4 < UAS system. Expression of GFP < Smo rescues the mutant phenotype of smo2 embryos (data not shown) and causes ectopic expression of dpp-lacZ in wing imaginal discs, indicating that the fusion protein is functional (Fig. 1a). We found that when expressed under the control of a ubiquitously active promoter the GFP < Smo protein accumulates preferentially in cells expressing or responding to Hh protein both in the embryo (not shown) and in the larval ectoderm (Fig. 1c,d). A similar modulation of GFP < Smo distribution is seen in wing imaginal discs cells in which the protein is expressed in the dorsal compartment of the disc under the control of the apterous-GAL4 driver (ap-GAL4). In this case, GFP < Smo accumulates preferentially in the posterior compartment and in cells just anterior to the A/P compartment boundary (Fig. 1a). Confocal analysis of the ectodermal cells of larvae expressing GFP < Smo reveals that most of the protein is localised to the plasma membrane (Fig. 1c,d). To analyse the effects of Ptc over-expression on Smo localisation, we generated a Yellow Fluorescent Protein (YFP) tagged form of Smo that behaves identically to the GFP < Smo. However, when expression of a Cyan Fluorescent Protein (CFP) tagged Ptc protein is driven simultaneously in the same cells, the YFP < Smo disappears from the plasma membrane and becomes predominantly localised intracellularly, accumulating in punctate vesicle-like structures, some of which also accumulate Ptc protein (Fig. 1e – h). To analyse the sub-cellular distribution of Smo at higher resolution, we next constructed a transgene containing the full-length Smo coding region fused in frame to the sequence encoding horseradish peroxidase (HRP). Expression of this construct in the wing imaginal disc causes the ectopic activation of dpp-lacZ, indicating that the fusion protein is functional. Visualisation of HRP activity by Diaminobenzidine (DAB) staining of imaginal discs in which the construct is expressed under ap-GAL4 control, revealed staining in both posterior and anterior compartments of the dorsal wing pouch, consistent with the well documented stability of HRP (data not shown). Transmission electron microscopy (TEM) of ultrathin sections of such stained imaginal discs, however, revealed quite distinct distributions of HRP activity within the anterior and posterior compartments of the disc (Fig. 2). Notably, cells in the posterior compartment show high levels of labelling in the plasma membrane whereas such labelling is essentially absent from anterior compartment cells. By contrast, anterior compartment cells show highest levels of labelling intracellularly in lysosomes, whereas in the posterior compartment, early and late endosomes along with a third small unidentified vesicle show significant
509
Fig. 2. Intracellular distribution of Smo < HRP in anterior and posterior compartment cells of the wing imaginal disc. Ultrathin sections of wing imaginal discs expressing Smo < HRP under the control of apterous < GAL4. Panel (a) shows cells from the dorsal anterior region of the wing blade; note the absence of signal from the plasma membrane and most organelles with the exception of very heavily stained vesicles that correspond to lysosomes. Panel (b) shows cells from the dorsal posterior region of a similar imaginal disc. In this case, the plasma membrane is highlighted by heavy staining as are several small, unidentified vesicles (arrowheads) and larger vesicles, corresponding to multivesicular bodies (arrows). Magnification in panel (a) is £ 9000. Magnification in panel (b) is £ 23000.
levels of labelling (Fig. 2). These findings are consistent with the observed distribution of the GFP < Smo protein described above and suggest that Smo is trafficked to the plasma membrane in Hh expressing or responding cells whilst being preferentially targeted to the degradative pathway in anterior compartment cells. 2.2. Smo protein levels are regulated by Ptc but not by Protein Kinase A (PKA) or Cos2 To confirm that the observed compartment specific accumulation of Smo is regulated by the activity of Ptc, we analysed the distribution of the endogenous Smo protein in wing discs containing clones of anterior compartment cells lacking Ptc activity. As expected, such clones exhibit
510
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 3. Neither PKA nor Costal2 activity is required to regulate Smo levels. Wing imaginal discs containing clones of anterior and posterior cells mutant for ptc (a –c), pka (d–f) and cos2 (g–l). Clones are marked by the absence of GFP (green). Smo protein distribution is shown in red. Anterior is to the left. Smo protein is upregulated in clones of A cells distant from the A/P boundary and mutant for ptc. However, Smo protein levels are unaffected in both anterior and posterior clones mutant for pka (d, f) or cos2 (g, i); note also that accumulation of Smo in anterior compartment cells near the A/P boundary is unaffected by the loss of Cos2 activity (j–l).
high levels of Smo protein accumulation relative to surrounding wild type cells (Fig. 3a – c). Since Ptc is a negative regulator of Hh signalling, it is possible that Smo levels might be stabilised by Hh signalling activity. In this case, activating Hh signalling by inactivating downstream negative regulators of the pathway ought to result in Smo stabilisation. Indeed it has been suggested that antagonising the activity of PKA, a negative regulator of the pathway, by over-expressing a dominant negative form of the protein, is sufficient to stabilise Smo in anterior compartment cells where it is usually downregulated (Alcedo et al., 2000). In contrast to this observation, we found that Smo levels are unaffected in clones of anterior cells mutant for PKA (Fig. 3d –f). It has also been suggested that Smo stability might be influenced by Cos-2 activity (Zhu et al., 2003; Hooper, 2003). However, we observed that clones of cells lacking Cos-2 activity showed no changes in the accumulation of the endogenous Smo protein (Fig. 3g – i). We note in particular that loss of Cos-2 activity led neither to an increase in Smo accumulation in Hh non-responding cells nor to a decrease in its accumulation in those cells that are responding to Hh at the A – P compartment boundary (Fig. 3j – l). Taken together, these data suggest that Smo protein is stabilised by Ptc inactivation rather than Hh signalling pathway activation and that Smo accumulation is a cause rather than a consequence of Hh signal transduction.
2.3. Activity of Smo gain of function mutations in imaginal discs Two oncogenic mutations in human Smo have been found in or close to the seventh transmembrane region (Xie et al., 1998) while several other constitutively active mutants have been selected on the basis of their resistance to the inhibitory effects of the alkaloid cyclopamine (Taipale et al., 2000). Experiments in cell culture have suggested that these proteins are ‘constitutively active’ because they escape Ptc repression (Xie et al., 1998). To investigate their effects in vivo, we introduced substitutions equivalent to either oncogenic mutant into the FLAG tagged smo cDNA (FLAG < SmoK580Q and FLAG < SmoW553L) and one of the cyclopamine resistant mutants into a cDNA encoding a GFP tagged variant of Smo (GFP < SmoA479Y) and used the GAL4-UAS system to misexpress each mutant cDNAs in the wing imaginal disc. By contrast to the wild type protein, the putative gain of function A479Y mutant form of GFP < Smo (GFP < SmoA479Y) accumulates at uniformly high levels in both compartments when expressed under ap-GAL4 control (Fig. 1b). Although expression of GFP < SmoA479Y by ap-GAL4 results in more extensive levels of ectopic dpp-lacZ upregulation (Fig. 1a,b), the resulting phenotype in the adult wing does not appear to be more severe than that
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
511
Fig. 4. The effects of Smo gain-of-function mutations in wing imaginal discs. Wings and wing imaginal discs in which FLAG-tagged Smo constructs have been misexpressed. (a) Ectopic expression of FLAG < Smo under the control of the 71B GAL4 driver results in ectopic veining around the second and third wing veins (L2 and L3), indicative of a mild activation of the Hh pathway. This can be seen more clearly in wing discs expressing FLAG < Smo under the control of 30A GAL4. The dpp-lacZ reporter gene is ectopically activated to a high level around the wing pouch (c), although no defect is observed in adult wings (b). Similar phenotypes are observed when oncogenic or constitutively active forms of Smo are misexpressed, although such cases exhibit a reduced level of ectopic dpp-lacZ expression compared with FLAG < Smo. (d–f) GFP < SmoA479Y; (g–i) FLAG < SmoK580Q; (j –l) FLAG < SmoW553L.
produced when GFP < Smo is expressed in the same way (data not shown). Misexpression of the FLAG tagged wild type protein driven by 30A causes ectopic activation of dpp-lacZ but has little if any effect on the differentiation of the wing; by contrast, misexpression using the 71B GAL4 (Brand and Perrimon, 1993) reproducibly results in the formation of extra veins in the anterior compartment of the wing blade (Fig. 4a). Surprisingly, we find that each of the putative gain of function mutant proteins, FLAG < SmoK580Q and FLAG < SmoW553L, behaves very similarly to the wild type protein when misexpressed in the same way; indeed, if anything they display a somewhat weaker activity, as judged both by the level of dpp-lacZ expression and the extent of extra venation formed when expressed under 71B control. 2.4. Smo function is disrupted by mutations in the amino-terminal and heptahelical domains Previous studies have identified several non-sense and two missense mutations in Smo; one of the latter causes a C–Y
substitution in the putative extra-cellular N-terminal domain (Chen and Struhl, 1998; Alcedo et al., 2000) whilst the other results in another C–Y substitution, in this case in the second transmembrane domain (Alcedo et al., 2000). To identify further critical residues in Smo, we undertook a screen for new EMS induced smo alleles (see Section 4). Amongst 8000 mutagenised second chromosomes screened, we identified seven new alleles on the basis of their lethality in trans to smo3 : Analysis of the cuticle patterns of homozygous embryos reveals a phenotypic series with smoIA3 ; smo2A and smo4DI exhibiting weaker phenotypes than smo3J ; smo1K ; smo4C2 and smo4F4 (Fig. 5). The first five of these alleles were characterised by single stranded conformation polymorphism (SSCP) analysis and/or DNA sequence analysis, the results of which are summarised schematically in Fig. 6. smo3J is associated with a stop codon at position 388 while smo1K has an 11 bp deletion after codon 437, resulting in a frameshift that cause premature termination after a further 15 codons. The three phenotypically weaker alleles by contrast, are all associated with missense mutations. In one of these, smo2A ; an Arg residue in the i3 loop, close to the boundary with TM6, is
512
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 5. Smo embryonic cuticle phenotypes. (a) Embryo homozygous for the smo2A allele showing a typical hypomorphic zygotic phenotype; note the partial fusion of abdominal denticle belts. (b) Embryo homozygous for the smo3 allele exhibiting a typical strong amorphic zygotic phenotype. (c) Partial rescue of the smo3 phenotype by prdGAL4 driven expression of FLAG < Smo; note the restoration of naked cuticle in alternate abdominal segments, corresponding to the domains of expression of the prdGAL4 driver.
substituted by a Cys. Notably, this Arg is conserved in all Frizzled and Smoothened proteins characterised to date. The smo4DI allele is associated with a Cys–Ser substitution at position 413 in the second putative extra-cellular loop, which is likely to disrupt the structure of the protein. The same substitution, this time in the Cys rich domain (CRD) of the putative extra-cellular N-terminal domain, is associated with the smoIA3 : While in vitro assays indicate that vertebrate Smo can retain activity in cultured mammalian cells in the absence of its CRD (Taipale et al., 2002), our finding that smoIA3 ; like smo2 (Alcedo et al., 2000) is associated with mutation of an Nterminal Cys residue, suggests that disulphide bridges within the CRD might play a critical role in effecting Smo activity in vivo. 2.5. The N-terminal extra-cellular domain is essential for Smo activity in vivo To investigate the requirements for the N-terminal extracellular domain (ECD) further, we generated DNFLAG <
Smo, a derivative of FLAG < Smo (Ingham et al., 2000) lacking all but 17 amino acids of the putative ECD. Ectopic expression of FLAG < Smo and of the DN derivative in wing imaginal discs using the GAL4 driver 30A results in uniformly high levels of FLAG immunoreactivity in both compartments, indicating that both transgenes are efficiently transcribed and translated throughout the enhancer trap domain (Fig. 7a,b). This contrasts with the post-translational modulation of endogenous Smo and of the GFP and HRP tagged forms described above. To confirm that the FLAG labelling accurately reflects the distribution of the transgene derived proteins, we also stained transgenic discs with an antibody raised against the C-terminus of Smo (Denef et al., 2000). This revealed an identical distribution in both cases (Fig. 7d, inset and data not shown), suggesting that the stability reflects the high levels of expression driven by the 30A GAL4 driver. Consistent with this, we found that GFP < Smo also accumulates at high levels in both compartments when expressed under 30A control (Fig. 7a, inset). Concomitant with the expression of FLAG, there is an increased accumulation of the full-length form of Ci and an ectopic activation of the dpp-lacZ reporter gene in anterior compartment cells expressing FLAG < Smo with 30A (Fig. 7b,c). Both of these effects are indicative of a derepression of the Hh pathway. By contrast, expression of DNFLAG < Smo in the same domain has no effect on Ci stability nor does it result in ectopic dpp-lacZ expression (Fig. 7d– f), confirming that the ECD is critical for Smo activity. 2.6. Progressive deletions reveal differential requirements for carboxy terminal regions of the Smo protein Since no mutations have been identified in the putative intracellular C-terminal domain (CTD) of Smo, we investigated the requirement for this region by deletion analysis. In the absence of any obvious motifs within CTD, we arbitrarily introduced stop codons at intervals of about 300 nucleotides, generating five different deletion constructs with progressively shorter intracellular tails. All carry the same N-terminal FLAG tag as the full-length FLAG < Smo. The largest of these, SmoDC0, ends immediately after the predicted seventh transmembrane region at residue T562 and thus lacks the entire CTD. Expression of this construct under 30A control results in high levels of FLAG accumulation but has no effect on dpplacZ expression (Fig. 8). SmoDC1 (DA638) similarly lacks any activity in this assay, indicating a requirement for the CTD that again contrasts with analysis of human Smo in tissue culture assays (Murone et al., 1999). By contrast, the three shorter deletion mutants, SmoDC2 (DT724), DC3 (DV837) and DC4 (DS939), all induce dpp-lacZ expression (Fig. 8). To test the activity of these constructs further, we performed rescue experiments on embryos mutant for both maternal and zygotic smo. Good restitution of ‘naked’
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
513
Fig. 6. Location of point mutations in the Smo protein. Schematic representation of the Smo protein showing its predicted heptahelical topology with long extra-cellular N-terminal, containing the CRD and intracellular CTD. Cys residues are indicated by green circles; mutated resides associated with different mutant alleles are indicated by red filled circles. The smoF5 and smo3 lesions were characterised by Chen and Struhl (1998) and the smo1 lesion by Alcedo et al. (2000); all other lesions shown were characterized in this study. Blue circles indicate the region of the C-terminal domain that is not conserved between the Drosophila and vertebrate forms of the protein. Smo1k not shown.
cuticle was seen with DC4, but already with DC3 rescue was variable and often incomplete (data not shown). Shorter constructs (three and two different lines were tested of DC1 and DC2, respectively) had no rescuing activity at all. Thus, this more stringent test revealed a requirement for the juxtamembrane half of the CTD, a region that has been highly conserved between Drosophila and vertebrates (see Quirk et al. (1997)).
3. Discussion The Smo protein is an essential component of the Hh signal transduction pathway; in all contexts analysed to date, inactivation of Smo renders cells incapable of responding to Hh signals (Lawrence et al., 1999b; Zhang and Kalderon, 2000; Zhang et al., 2001; Chen et al., 2001). Reciprocally, gain of function mutations have been isolated in human Smo that are sufficient to activate the Hh pathway
in the absence of ligand (Xie et al., 1998; Taipale et al., 2000; Hynes et al., 2000). Hh signalling regulates Smo activity via the multipass transmembrane receptor protein Ptc. In the absence of Hh ligand, Ptc represses Smo activity whereas when Hh binds to Ptc, Smo becomes active. Immunoprecipitation studies of vertebrate Smo from tissue culture cells co-expressing Ptc and Smo have suggested that this Ptc-dependent modulation of Smo activity is mediated by a direct interaction between the two proteins, a view consistent with the effects of high level Smo expression in cultured cells carrying a Shhresponsive reporter gene (Murone et al., 1999). Studies in Drosophila, however, have suggested that Ptc suppresses Smo activity by regulating the sub-cellular distribution and stability of the protein in a non-stoichiometric manner (Alcedo et al., 2000; Denef et al., 2000; Ingham et al., 2000; Taipale et al., 2002). The relatively weak effects of ectopic Smo expression in the wing imaginal disc described here and noted previously by other authors (Denef et al., 2000;
514
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 7. Effects of misexpression of FLAG < Smo and its N-terminally truncated derivative in wing imaginal discs. (a –c) Expression of FLAG < Smo under 30A control (a) results in ectopic accumulation of Ci155 (b) and ectopic activation of dpp-lacZ (c) around the edge of the presumptive wing blade in the anterior compartment, corresponding to those cells in which the transgene is expressed. (d–f) Similar 30A driven expression of DNFLAG < Smo (d) has no effect on Ci155 stability (e) or dpp-lacZ expression (f). (a, inset) Expression of GFP < Smo with 30A GAL4 shows uniform levels of GFP expression in both anterior and posterior compartments (compare with Fig. 6a) indicative of the strength of the 30A driver. (d, inset) Expression of DNFLAG:Smo detected with antibodies against the C-terminal domain of Smo. High levels of Smo expression in the anterior compartment demonstrate stable Smo expression from the transgene, even in the absence of the N-terminus, and are consistent with the strength of the 30A driver.
Martin et al., 2001; Zhu et al., 2003), presumably reflect this sub-stoichiometric regulation of the exogenously supplied protein by endogenous Ptc activity. Consistent with this, we find that ectopically expressed Smo protein is clearly subject to Ptc-mediated destabilisation, as revealed by the distribution of the GFP and YFP tagged forms of Smo. Nevertheless, when expressed at very high levels, as occurs for instance using the 30A GAL4 driver line, the Smo protein can escape the effects of Ptc activity and accumulate in anterior compartment cells where it activates Hh target gene activity. Even under these circumstances, however, the phenotypic effects of its over-expression are relatively mild. Smo shares limited sequence homology with members of the Frizzled family of Wnt receptors, in particular in the CRD of the putative N-terminal ECD that in Fz has been shown to be necessary and sufficient for Wnt binding (Bhanot et al., 1996). By contrast with Fz, there is no evidence that the ECD of Smo binds Wg (or other Wnts) or indeed Hh, consistent with the finding that Smo activity is independent of Hh in the absence of the inhibitory effects
of Ptc. Nevertheless, of the five Smo point mutants that we have identified, three affect residues in the N-terminal ECD, including two Cys residues that are highly conserved in all Fz family members, suggesting ECD and the CRD in particular plays some critical role in Smo function (see also Chen and Struhl (1998)). In line with this, we find that deletion of the ECD results in a loss of function of the protein when assayed both in our mutant rescue and imaginal disc ectopic expression systems. One explanation for this requirement could be that the N-terminus binds an as yet unidentified ligand that is required even in the absence of Ptc repressive activity, or indeed the secretion of which is inhibited by Ptc. Alternatively, the ECD may itself function as an activating ligand (see Chen and Struhl (1998)). The protease-activated G-protein coupled receptors (PARs) provide a precedent for such intramolecular receptor activation (Nystedt et al., 1994). PARs are activated by proteolytic cleavage of the ECD, the tip of the new amino terminus created by this cleavage acting as a tethered ligand by binding to the receptor core. We emphasise, however, that there is no evidence that Smo undergoes
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
515
Fig. 8. Progressive deletion of the C-terminal domain inactivates FLAG:Smo. The upper panels show the high level expression of transgenes encoding the C-terminal deletion derivatives of FLAG:Smo under 30A control, as revealed by anti-FLAG staining (green). Lower panels show the expression of dpp-lacZ (red) in discs expressing the respective transgenes.
a similar proteolytic activation. In contrast to our findings, Murone et al. (1999) have previously demonstrated that an N-terminally deleted form of human Smo retains activity when over expressed in 10T1/2 cells. The disparity between this result and our own findings may reflect differences in the levels of over-expression achieved in the in vivo versus the in vitro systems; it may be that at high enough levels, the mutant form of Smo can saturate the repressive activity of endogenous Ptc in the 10T1/2 cells, thus leading to the activation of the endogenous wild type Smo protein. Since no mutations have been identified in the CTD of Smo, we have also investigated the requirement for this region by deletion analysis. In contrast to the results of in vitro studies of human Smo suggesting that the terminal domain is dispensable (Murone et al., 1999) we found that deletion of the entire domain or even 80% of it completely inactivates the protein. Only deletions that preserve the juxtamembrane half of the C-terminal tail retain any activity; notably it is this part of the C-terminus that has been highly conserved between Drosophila and vertebrates (Quirk et al., 1997). Although these results demonstrate a requirement for carboxy terminal regions of the protein, they do not give any direct indications as to their molecular function. For example, structures in the carboxy terminus could be needed to keep other parts of Smo in the correct conformation. Alternatively, the carboxy terminus could bind directly to another protein or proteins that transduce
the Hh signal to the multimeric protein complex that regulates Ci stability and activation, or indeed to the multimeric complex itself. Amongst the new smo mutant alleles that we have isolated, we found one, smo2A ; to be a missense mutation that converts a highly conserved arginine to a cysteine at amino acid position 474 in the third intracellular loop. Residues at the C-terminal end of i3 in GPCRs have been shown to be critical for appropriate coupling to heterotrimeric G-proteins (Zhang et al., 2000) and loss of charged residues in this region of the angiotensin II receptor type I abolishes such coupling (Ohyama et al., 1992). While it has been shown that Frizzled proteins can transduce Wnt signals via heterotrimeric G-proteins (Malbon et al., 2001), the evidence for G-protein involvement in Smo activity remain equivocal (DeCamp et al., 2000; Murone et al., 1999; Norris et al., 2000). Both smo4DI and smo2A hypomorphic mutations reside in the heptahelical domain of the protein, the structural integrity of which has been implicated in Smo activity. Small molecule agonists and antagonists of vertebrate Smo have been shown to bind to this domain, implying its involvement in regulating the activity of the protein (Chen et al., 2002a,b). Exactly how these small molecules influence Smo activity is unclear, but one suggestion is that they may alter the conformation of the protein, respectively, towards a more active or inactive state (Chen et al., 2002b). In this view, the normal regulation of Smo by Ptc might involve the transport of a cellular small
516
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
molecule agonist/antagonist that similarly binds to Smo, altering its conformation. Alternatively, Ptc might act to redistribute Smo to a membrane compartment where it is susceptible to such an antagonist ligand or inaccessible to an agonist. Our analysis of the sub-cellular distribution of tagged forms of the Smo protein in Drosophila tissues is consistent with a role for Ptc in the intracellular trafficking of Smo (Martin et al., 2001; Strutt et al., 2001; Incardona et al., 2002). Specifically we find that Smo accumulates in the plasma membrane of cells that lack Ptc activity, whereas when co-expressed with Ptc, the same protein localises exclusively intracellularly, accumulating in distinct punctate structures. Similar results using the salivary gland as an assay system have recently been reported (Zhu et al., 2003). Significantly, our electron microscopy analysis of the HRP < Smo fusion proteins reveals high levels of HRP activity in the lysosomes of cells in which Ptc is active whereas in Ptc negative cells, the protein accumulates predominantly in the plasma membrane as well as in endosomes and unidentified presumptive transport vesicles. Analysis of the localisation and activity of putative hypermorphic forms of Smo might afford an insight into the mechanism of Smo function. Interestingly, when driven with ap-GAL4, the putative hypermorphic form of Smo, GFP < SmoA479, accumulates in the plasma membrane irrespective of Ptc activity and yields a stronger phenotype than GFP < Smo expressed under the same conditions, as assayed by dpp-lacZ expression. Surprisingly, however, the resulting adult phenotype appears to be no more severe in the case of the putative hypermorph than the wild type form. Furthermore, the over-expression of SmoA479Y as well as two other putative gain of function forms of Smo, SmoK580Q and SmoW553L, directed by the 30A and 71B GAL4 drivers resulted in lower levels of ectopic dpp-lacZ activation than those induced by wild type Smo. Consistent with this, the resulting adult phenotypes were also weaker in the case of the mutant forms of Smo than their wild type counterpart. We note, however, that Zhu et al. (2003) reported that the over-expression of certain putative hypermorphic forms of Smo results in stronger phenotypes than wild type Smo. The reason for these disparities is currently unclear. Finally, our observations that the cellular distribution of Smo is unaffected in imaginal disc clones lacking the activity of PKA or Cos2 (and hence constitutively transducing the Hh signal) has suggested Hh signalling is not sufficient to stabilise Smo and that Smo stabilisation is therefore more likely to be a cause, rather than an effect, of Hh signal transduction. However, Zhu et al. (2003) observed that over-expression of Cos2 in salivary gland cells can result in Smo destabilisation, suggesting that Cos2 activity is capable of influencing Smo levels. Taken together with our data, this suggests that antagonising Cos2 activity might be a necessary, but not sufficient, prerequisite for Smo stabilisation at the cell surface.
In conclusion, our data favour a mechanism whereby Ptc regulates Smo activity through inhibition of its accumulation in the plasma membrane, targeting it instead to the lysosomal pathway. In this view, Ptc might regulate Smo activity simply by modulating the levels of protein present in the cell. Alternatively, it may be that the sub-cellular location of Smo is critical for its activation, the plasma membrane perhaps providing an environment in which Smo is accessible to an intracellular agonist.
4. Materials and methods In vivo mutagenesis—cn bw sp males were mutagenised with ethylmethylsulfonylurea (EMS) using standard protocols (Ingham, 1984). Eight thousand chromosomes were screened for lethality in trans to the smo3 allele. The embryonic phenotypes of all non-complementing mutations isolated in this way were tested in trans to smo1 and smo2 alleles. In vitro mutagenesis—DNFlagsmo: A smo cDNA missing the sequence for the extra-cellular amino terminus was amplified using the sense primer 50 -TTATGCATTGC GATGCAAGGATCCACTC-30 , adding an Nsi I site, and 50 -TTCTCGAGCTATTTTGAAGGCAGCAATAACATT TTGAG-30 , adding an Xho I site after the translation stop codon. The fragment was cut with Nsi I and Xho I and cloned between the same sites in pSKII/SpFlag. FlagsmoDC constructs: A series of 30 deletion mutants of the full-length Flag-tagged smo cDNA (Ingham et al., 2000) were made using the following antisense primers (all containing a stop codon before an Xho I site): DT562: 50 -TTCTCGAGCTAAGTCTCAATTGAAGAA GGTG-30 ; DA638: 50 -TTCTCGAGCTATGCAGCCCAA GTTGATGA-30 ; DT742: 50 -TTCTCGAGCTAAGTACTG CTTTCCCTTCTCC-30 ; DV837: 50 -TTCTCGAGCTAT ACAACCACGTTAAGATCCTG-30 ; DS939: 5 0 -TTCT CGAGCTAGCTGTTAAGAAGGCTTCCTAC-30 , paired with suitable sense primers. Amplified fragments were digested with Xho I and Spe I or EcoR I, recognising internal sites, and cloned in pSKII/Flagsmo that had been cut with the same enzymes. UAS-Smo-EGFP construct: the stop codon of the smo cDNA was replaced by a KpnI site using standard PCR-based cloning. The resulting full-length smo cDNA with KpnI site at the C-terminus was cloned into pEGFP-N1 vector (Clontech). These constructs were introduced in w1118 host flies by P element-mediated transformation using standard methods. UAS-Boss-HRP-Smo construct: the signal sequences and HRP sequences were obtained from the Boss-HRP construct (Sunio et al., 1999) and 30 to the DNA encoding the signal sequence of Smo from glycine 37 was ligated using standard PCR-based cloning. The resulting sequence was M(1)…CH(33)agMQLT…NSGGklG(37)STTP. Boss signal sequences are in bold; the extra amino acids
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
generated by the fusion are indicated in lower case. The HRP sequences are italicised and bold letters indicate the amino acids from Smoothened. Smo gain of function mutants: point mutations resulting in amino acid substitutions were introduced following the manufacturer’s protocol. Smo cDNA was subcloned into a p-EGFP-N1 vector (clontech) in order to generate a GFP fusion protein with EGFP at the C-terminus (smo-EGFP). Smo cDNA was cut with the enzyme BclI to generate a fragment that was subcloned into bluescript pSKII. A point mutation (A479Y) was introduced by site directed mutagenesis (using Stratagene Quickchange kit). The mutated BclI fragment was then re-introduced into the smo-EGFP to make smoA479Y-EGFP. The smoA479YEGFP cDNA was released from the p-EGFP-N1 vector using NotI. All modified smo cDNAs were cloned into pUAST vector (Brand and Perrimon, 1993). The proofreading polymerase Pfu (Stratagene) was used throughout these manipulations and all new fragments were sequenced to check for any errors. Transgenic flies—A mixture of expression construct plasmid (< 450 ng/ml) and the transpose-encoding plasmid pUChsD2 –3 (< 150 ng/ml; plasmid kindly provided by Dr David I. Strutt) in water was injected into blastoderm yw embryos using a Leitz micromanipulator fitted to a Leica DM IL inverted microscope. Surviving adult flies were crossed to yw flies and transformant progeny identified by their pigmented eyes. Insertions were mapped to chromosome and balanced. Stocks were kept at 18 8C and crosses were done at 25 8C, unless otherwise indicated. Clonal analysis—First instar larvae were heat-shocked for 2 h at 38 8C on two consecutive days. Imaginal discs were dissected from wandering third instar larvae, fixed for 25 min in 4% paraformaldehyde in PBS and stained with antibodies to Smo (S. Cohen) Genotypes of the larvae were as follows: ptc clones: w hs-flp; FRT42D ptcS2/FRT42D ubi-GFP pka clones: w hs-flp; pka-c1 H2 FRT40A/ubi-GFP FRT40A cos2 clones: w hs-flp; FRT42D cos2k16101/FRT42D ubi-GFP Cuticle analysis—Unhatched first instar larvae were collected from apple juice plates, dechorionated in bleach and mounted directly in a 1:1 mixture of Hoyer’s medium and lactic acid. Wings from adult flies were dissected in isopropanol and mounted on microscope slides in Euparal. Antibodies—Monoclonal a-Flag antibodies M1 and M2 (Sigma) were used at a dilution of 1:200. M1 recognises an amino-terminal Flag tag sequence, providing a good indication that the transgenic protein has been correctly processed and inserted in the membrane. Monoclonal antiCi antibody 2A1 (kindly provided by Robert Holmgren) was used at 1:5 dilution. Monoclonal a-HA antibody 12CA5 (Boehringer) was used at 1:100 dilution. Rat polyclonal
517
anti-Smo antibody (kindly provided by Stephen Cohen) was used at 1:500. Rabbit a-b galactosidase antiserum was from Cappell and used at a dilution of 1:200. Fluorescent secondary antibodies were from Jackson Immunoresearch Laboratories Inc. or from Molecular Probes. Microscopy and image analysis—Images of fluorescently labelled wing discs were captured using a Leica TCS NT confocal microscope. The full depth of the discs was scanned in series of eight sections, which were then averaged using the ‘extended focus’ software function. Images were manipulated and the final figures composed using Adobe Photoshop CS software on Macintosh computers. 4.1. Transmission electron microscopy Embryos were fixed 10 min in heptane saturated with glutaraldehyde and devitellinised by hand in PBS with Tween (PTW). HRP activity was detected by incubation in DAB (0.5 mg/ml) with 0.003% H2O2 for 30 min. Embryos were washed with PBS and then processed for EM. Wing imaginal discs were dissected in PTW and incubated in DAB (0.5 mg/ml) with 0.003% H2O2 for 20 min. After washing in PTW, wing imaginal discs were fixed in 2% glutaraldehyde for 30 min and then washed in PBS and processed for EM. For EM, samples were embedded as described (Van Vactor et al., 1991). Ultrathin sections, approximately 70 – 90 nm thick, were post-stained with Reynold’s Lead Citrate and examined using a Philip CM10 Transmission Electron Microscope.
Acknowledgements We thank Bill Brown, Liz Jones, Debbie Sutton and Jeremy Quirk for help with the SSCP and sequence analysis and John Proctor for his expert assistance with the TEM. This work was supported by grants from Ontogeny inc. and the Wellcome Trust to PWI. AAS is funded by an MRC postgraduate studentship. The Centre for Developmental Genetics confocal microscopy facility is funded by Yorkshire Cancer Research.
References Alcedo, J., Ayzenzon, M., Vonohlen, T., Noll, M., Hooper, J., 1996. The Drosophila Smoothened gene encodes a 7-pass membrane-protein, a putative receptor for the Hedgehog signal. Cell 86, 221–232. Alcedo, J., Zou, Y., Noll, M., 2000. Post-transcriptional regulation of Smoothened is part of a self-correcting mechanism in the Hedgehog signaling system. Mol. Cell 6, 457 –465. Bhanot, P., Brink, M., Samos, C.H., Hsieh, J.-C., Wang, Y., Macke, J.P., et al., 1996. A new member of the Frizzled family from Drosophila functions as a Wingless receptor. Nature 382, 225–230. Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401–415.
518
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Chen, Y., Struhl, G., 1998. In vivo evidence that Patched and Smoothened constitute distinct binding and transducing components of a Hedgehog receptor complex. Development 125, 4943–4948. Chen, W., Burgess, S., Hopkins, N., 2001. Analysis of the zebrafish Smoothened mutant reveals conserved and divergent functions of Hedgehog activity. Development 128, 2385–2396. Chen, J., Taipale, J., Cooper, M., Beachy, P., 2002a. Inhibition of Hedgehog signaling by direct binding of cyclopamine to Smoothened. Genes Dev. 16, 2743– 2748. Chen, J., Taipale, J., Young, K., Maiti, T., Beachy, P., 2002b. Small molecule modulation of Smoothened activity. Proc. Natl. Acad. Sci. USA 99, 14071–14076. DeCamp, D., Thompson, T., de Sauvage, F., Lerner, M., 2000. Smoothened activates Ga i-mediated signaling in frog melanophores. J. Biol. Chem. 275, 26322–26327. Denef, N., Neubuser, D., Perez, L., Cohen, S., 2000. Hedgehog induces opposite changes in turnover and subcellular localization of Patched and Smoothened. Cell 102, 521–531. van den Heuvel, M., Ingham, P.W., 1996. Smoothened encodes a serpentine protein required for Hedgehog signaling. Nature 382, 547 –551. Hooper, J.E., 2003. Smoothened translates Hedgehog levels into distinct responses. Development 130, 3951–3963. Hynes, M., Ye, W., Wang, K., Stone, D., Murone, M., Sauvage, F., Rosenthal, A., 2000. The seven-transmembrane receptor Smoothened cell-autonomously induces multiple ventral cell types. Nat. Neurosci. 3, 41 –46. Incardona, J., Gruenberg, J., Roelink, H., 2002. Sonic Hedgehog induces the segregation of Patched and Smoothened in endosomes. Curr. Biol. 12, 983 –995. Ingham, P.W., 1984. A gene that regulates the Bithorax complex differentially in larval and adult cells of Drosophila. Cell 37, 815–823. Ingham, P.W., McMahon, A.P., 2001. Hedgehog signaling in animal development: paradigms and principles. Genes Dev. 15, 3059–3087. Ingham, P., Nystedt, S., Nakano, Y., Brown, W., Stark, D., van den Heuvel, M., Taylor, A., 2000. Patched represses the Hedgehog signalling pathway by promoting modification of the Smoothened protein. Curr. Biol. 10, 1315–1318. Lawrence, P., Casal, J., Struhl, G., 1999b. The Hedgehog morphogen and gradients of cell affinity in the abdomen of Drosophila. Development 126, 2441–2449. Malbon, C., Wang, H., Moon, R., 2001. Wnt signaling and heterotrimeric G-proteins: strange bedfellows or a classic romance? Biochem. Biophys. Res. Commun. 287, 589–593. Martin, V., Carrillo, G., Torroja, C., Guerrero, I., 2001. The sterol-sensing domain of Patched protein seems to control Smoothened activity through Patched vesicular trafficking. Curr. Biol. 11, 601 –607. McMahon, A.P., Ingham, P.W., Tabin, C., 2003. The developmental roles and clinical significance of Hedeghog signalling. Curr. Topics Dev. Biol. 53, 1–114. Murone, M., Rosenthal, A., de Sauvage, F., 1999. Sonic Hedgehog signaling by the Patched–Smoothened receptor complex. Curr. Biol. 9, 76–84.
Norris, W., Neyt, C., Ingham, P., Currie, P., 2000. Slow muscle induction by Hedgehog signalling in vitro. J. Cell Sci. 113, 2695–2703. Nystedt, S., Emilsson, K., Wahlestedt, C., Sundelin, J., 1994. Molecular cloning of a potential proteinase activated receptor. Proc. Natl. Acad. Sci. USA 91, 9208–9212. Ohyama, K., Yamano, Y., Chaki, S., Kondo, T., Inagami, T., 1992. Domains for G-protein coupling in angiotensin II receptor type I: studies by site-directed mutagenesis. Biochem. Biophys. Res. Commun. 189, 677–683. Quirk, J., van den Heuvel, M., Henrique, D., Marigo, V., Jones, T., Tabin, C., Ingham, P., 1997. The Smoothened gene and Hedgehog signal transduction in Drosophila and vertebrate development. Cold Spring Harbor Symp. 62, 217 –226. Strutt, H., Thomas, C., Nakano, Y., Stark, D., Neave, B., Taylor, A., Ingham, P., 2001. Mutations in the sterol sensing domain of Patched suggest a role for vesicular trafficking in Smoothened regulation. Curr. Biol. 11, 608 –613. Sunio, A., Metcalf, A., Kra¨mer, H., 1999. Genetic dissection of endocytic trafficking in Drosophila using a horseradish peroxidase-bride of sevenless chimera: hook is required for normal maturation of multivesicular endosomes. Mol. Biol. Cell 10, 847 –859. Taipale, J., Chen, J., Cooper, M., Wang, B., Mann, R., Milenkovic, L., et al., 2000. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 406, 1005–1009. Taipale, J., Cooper, M.K., Maiti, T., Beachy, P.A., 2002. Patched acts catalytically to suppress the activity of Smoothened. Nature 418, 892 –897. Van Vactor, D.L. Jr., Cagan, R., Kra¨mer, H., Zipursky, S., 1991. Induction in the developing compound eye of Drosophila: multiple mechanisms restrict R7 induction to a single retinal precursor cell. Cell 67, 1145–1155. Varga, Z.M., Amores, A., Lewis, K.E., Yan, Y.L., Postlethwait, J.H., Eisen, J.S., Westerfield, M., 2001. Zebrafish Smoothened functions in ventral neural tube specification and axon tract formation. Development 128, 3497–3509. Xie, J., Murone, M., Luoh, S., Ryan, A., Gu, Q., Zhang, C., et al., 1998. Activating Smoothened mutations in sporadic basal-cell carcinoma. Nature 391, 90–92. Zhang, Y., Kalderon, D., 2000. Regulation of cell proliferation and patterning in Drosophila oogenesis by Hedgehog signaling. Development 127, 2165–2176. Zhang, M., Zhao, X., Chen, H., Catt, K., Hunyady, L., 2000. Activation of the AT1 angiotensin receptor is dependent on adjacent apolar residues in the carboxyl terminus of the third cytoplasmic loop. J. Biol. Chem. 275, 15782–15788. Zhang, X., Ramalho-Santos, M., McMahon, A., 2001. Smoothened mutants reveal redundant roles for Shh and Ihh signaling including regulation of L/R asymmetry by the mouse node. Cell 105, 781 –792. Zhu, A., Zheng, L., Suyama, K., Scott, M., 2003. Altered localization of Drosophila Smoothened protein activates Hedgehog signal transduction. Genes Dev. 17, 1240–1252.
Functional domains and sub-cellular distribution of the Hedgehog transducing protein Smoothened in Drosophila Y. Nakano1,2, S. Nystedt1, A.A. Shivdasani1, H. Strutt1, C. Thomas1, P.W. Ingham* MRC Intercellular Signalling Group, Centre for Developmental Genetics, University School of Medicine and Biomedical Science, Firth Court, Western Bank, Sheffield S10 2TN, UK Received 17 September 2003; received in revised form 30 March 2004; accepted 15 April 2004 Available online 28 April 2004
Abstract The Hedgehog signalling pathway is deployed repeatedly during normal animal development and its inappropriate activity is associated with various tumours in human. The serpentine protein Smoothened (Smo) is essential for cells to respond to the Hedeghog (Hh) signal; oncogenic forms of Smo have been isolated from human basal cell carcinomas. Despite similarities with ligand binding G-protein coupled receptors, the molecular basis of Smo activity and its regulation remains unclear. In non-responding cells, Smo is suppressed by the activity of another multipass membrane spanning protein Ptc, which acts as the Hh receptor. In Drosophila, binding of Hh to Ptc has been shown to cause an accumulation of phosphorylated Smo protein and a concomitant stabilisation of the activated form of the Ci transcription factor. Here, we identify domains essential for Smo activity and investigate the sub-cellular distribution of the wild type protein in vivo. We find that deletion of the amino terminus and the juxtamembrane region of the carboxy terminus of the protein result in the loss of normal Smo activity. Using Green Fluorescent Protein (GFP) and horseradish peroxidase fusion proteins we show that Smo accumulates in the plasma membrane of cells in which Ptc activity is abrogated by Hh but is targeted to the degradative pathway in cells where Ptc is active. We further demonstrate that Smo accumulation is likely to be a cause, rather than a consequence, of Hh signal transduction. q 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Smoothened; Oncogene; Signalling; Patched; Hedgehog; Drosophila
1. Introduction Despite substantial knowledge of the plethora of processes in which Hedeghog (Hh) signals are deployed during animal development and the growing recognition of the role played by aberrant Hh signalling in various human tumours (see McMahon et al. (2003)), many gaps remain in our understanding of how the Hh signal is transduced. In particular, the mechanisms by which the Smoothened (Smo) protein—an obligate component of the pathway (Alcedo et al., 1996; van den Heuvel and Ingham, 1996; Chen et al., 2001; Varga et al., 2001; Zhang et al., 2001) and a target for oncogenic mutation (Xie et al., 1998)—is regulated and activates its intracellular targets are still largely obscure. Although it shares structural similarities with serpentine receptor proteins, particularly members of the Frizzled * Corresponding author. Tel.: þ44-114-222-2803; fax: þ 44-114-2222788. E-mail address: [email protected] (P.W. Ingham). 1 These authors made equal contributions and are listed in alphabetical order. 2 Present address: Department of Genetics, Hyogo College of Medicine, Mukogawa-cho, 1-1, Nishinomiya 663-8501, Japan. 0925-4773/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mod.2004.04.015
family of Wnt receptors (Quirk et al., 1997) genetic and biochemical data indicate that Smo is not directly regulated by Hh ligand. Instead, its activity is somehow controlled by the multipass transmembrane protein Patched (Ptc), which itself acts as the Hh receptor. In unstimulated cells Ptc represses Smo activity; binding of Hh to Ptc relieves this repression, allowing Smo to signal to a multimeric complex inside the cell that includes the transcription factor Ci. In the absence of Hh signalling, Ci is sequestered in a cytoplasmic complex that includes the kinesin related protein Cos-2 and the serine threonine kinase Fused. This sequestration inhibits nuclear import of the full-length 155 kDa protein and promotes its cleavage to a 75 kDa form that can enter the nucleus and repress transcription. Activation of Smo inhibits Ci cleavage and activates the full-length protein, resulting in the transcription of Hh-responsive target genes (reviewed in Ingham and McMahon (2001)). The way in which the activity of Smo is transmitted to the Ci-Cos-2-Fu complex is obscure. Whilst Smo has a predicted heptahelical domain (Alcedo et al., 1996; van den Heuvel and Ingham, 1996), the signature motif of G-protein coupled receptors (GPCRs), there is no reported primary sequence similarity with any bona fide GPCR.
508
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Activity of the vertebrate Smo protein can be modulated by the alkaloid cyclopamine and other small molecule antagonists and agonists that have been shown to bind to the heptahelical domain (Chen et al., 2002a,b). Exactly how these small molecules influence Smo activity is unclear, but one suggestion is that they may alter the conformation of the protein, respectively, towards a more active or inactive state (Chen et al., 2002b). In this view, the normal regulation of Smo by Ptc might involve the transport of a cellular small molecule agonist/antagonist that similarly binds to Smo, altering its conformation. In line with this scenario, Ptc has been shown to share essential structural motifs with a family of bacterial permeases. Mutations in these motifs lead to a loss of its Smo suppressing activity (Taipale et al., 2002). Such permease activity would also be consistent with a role for Ptc in redistribution of Smo to a membrane compartment where it is susceptible to such an antagonist ligand or inaccessible to an agonist. In this view, the accumulation of Smo in the appropriate cellular location would be the key determinant of Smo activation.
Here, we present the results of in vivo analyses in Drosophila designed to investigate the functional domains and sub-cellular distribution of the Smo protein. Through both in vivo and in vitro mutagenesis, we have identified regions of the protein critical for its function. Using fusion proteins we have studied the changes in sub-cellular distribution of Smo in response to Hh and Ptc activity in vivo. Our results reveal critical requirements for both the amino and carboxy terminal domains for normal Smo activity and suggest that Ptc regulates Smo activity at least in part by targeting it to the lysosome.
2. Results 2.1. The sub-cellular distribution of Smo is regulated by Hh and Ptc activity Previous studies have shown that Smo protein accumulates preferentially in Hh responding cells in the Drosophila
Fig. 1. Regulation of Smo distribution in the wing imaginal discs and larval ectoderm. A GFP-tagged form of Smo expressed in wing imaginal discs under the control of the ap-GAL4 driver, accumulates in cells of the posterior compartment and just anterior to the boundary (right) but becomes degraded in more anterior cells (left) that do not receive the Hh signal (a, a0 ). By contrast, the distribution of GFP < SmoA479Y, a putative gain of function, tagged Smo mutant, does not appear to be regulated; the protein accumulating equally in anterior and posterior compartments (b, b0 ); the expression of dpp-lacZ (red) marks the anterior/posterior compartment boundary and ectopic Hh signal transduction in the anterior compartment; in both (a) and (b) dorsal is up. (c–e) Fluorescently tagged Smo molecules are post-transcriptionally regulated in larval ectoderm cells when expressed ubiquitously with act-GAL4 (top panel). High magnification projected images of several confocal sections (d, e) indicate that the tagged Smo accumulates predominantly at the plasma membrane where it appears to aggregate into clusters. (f–h) Simultaneous misexpression of both YFP < Smo and CFP < Ptc, results in the loss of Smo from the plasma membrane into intracellular vesicles. Occasionally, co-localisation of CFP < Ptc and YFP < Smo can be observed (arrow).
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
embryo and imaginal discs; moreover, in clone 8 cells the proportion of Smo present at the cell surface increases on exposure to Hh protein (Denef et al., 2000). To analyse the modulation of Smo distribution by Hh in vivo we generated a transgene encoding a GFP tagged form of the protein and expressed this in transgenic animals using the GAL4 < UAS system. Expression of GFP < Smo rescues the mutant phenotype of smo2 embryos (data not shown) and causes ectopic expression of dpp-lacZ in wing imaginal discs, indicating that the fusion protein is functional (Fig. 1a). We found that when expressed under the control of a ubiquitously active promoter the GFP < Smo protein accumulates preferentially in cells expressing or responding to Hh protein both in the embryo (not shown) and in the larval ectoderm (Fig. 1c,d). A similar modulation of GFP < Smo distribution is seen in wing imaginal discs cells in which the protein is expressed in the dorsal compartment of the disc under the control of the apterous-GAL4 driver (ap-GAL4). In this case, GFP < Smo accumulates preferentially in the posterior compartment and in cells just anterior to the A/P compartment boundary (Fig. 1a). Confocal analysis of the ectodermal cells of larvae expressing GFP < Smo reveals that most of the protein is localised to the plasma membrane (Fig. 1c,d). To analyse the effects of Ptc over-expression on Smo localisation, we generated a Yellow Fluorescent Protein (YFP) tagged form of Smo that behaves identically to the GFP < Smo. However, when expression of a Cyan Fluorescent Protein (CFP) tagged Ptc protein is driven simultaneously in the same cells, the YFP < Smo disappears from the plasma membrane and becomes predominantly localised intracellularly, accumulating in punctate vesicle-like structures, some of which also accumulate Ptc protein (Fig. 1e – h). To analyse the sub-cellular distribution of Smo at higher resolution, we next constructed a transgene containing the full-length Smo coding region fused in frame to the sequence encoding horseradish peroxidase (HRP). Expression of this construct in the wing imaginal disc causes the ectopic activation of dpp-lacZ, indicating that the fusion protein is functional. Visualisation of HRP activity by Diaminobenzidine (DAB) staining of imaginal discs in which the construct is expressed under ap-GAL4 control, revealed staining in both posterior and anterior compartments of the dorsal wing pouch, consistent with the well documented stability of HRP (data not shown). Transmission electron microscopy (TEM) of ultrathin sections of such stained imaginal discs, however, revealed quite distinct distributions of HRP activity within the anterior and posterior compartments of the disc (Fig. 2). Notably, cells in the posterior compartment show high levels of labelling in the plasma membrane whereas such labelling is essentially absent from anterior compartment cells. By contrast, anterior compartment cells show highest levels of labelling intracellularly in lysosomes, whereas in the posterior compartment, early and late endosomes along with a third small unidentified vesicle show significant
509
Fig. 2. Intracellular distribution of Smo < HRP in anterior and posterior compartment cells of the wing imaginal disc. Ultrathin sections of wing imaginal discs expressing Smo < HRP under the control of apterous < GAL4. Panel (a) shows cells from the dorsal anterior region of the wing blade; note the absence of signal from the plasma membrane and most organelles with the exception of very heavily stained vesicles that correspond to lysosomes. Panel (b) shows cells from the dorsal posterior region of a similar imaginal disc. In this case, the plasma membrane is highlighted by heavy staining as are several small, unidentified vesicles (arrowheads) and larger vesicles, corresponding to multivesicular bodies (arrows). Magnification in panel (a) is £ 9000. Magnification in panel (b) is £ 23000.
levels of labelling (Fig. 2). These findings are consistent with the observed distribution of the GFP < Smo protein described above and suggest that Smo is trafficked to the plasma membrane in Hh expressing or responding cells whilst being preferentially targeted to the degradative pathway in anterior compartment cells. 2.2. Smo protein levels are regulated by Ptc but not by Protein Kinase A (PKA) or Cos2 To confirm that the observed compartment specific accumulation of Smo is regulated by the activity of Ptc, we analysed the distribution of the endogenous Smo protein in wing discs containing clones of anterior compartment cells lacking Ptc activity. As expected, such clones exhibit
510
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 3. Neither PKA nor Costal2 activity is required to regulate Smo levels. Wing imaginal discs containing clones of anterior and posterior cells mutant for ptc (a –c), pka (d–f) and cos2 (g–l). Clones are marked by the absence of GFP (green). Smo protein distribution is shown in red. Anterior is to the left. Smo protein is upregulated in clones of A cells distant from the A/P boundary and mutant for ptc. However, Smo protein levels are unaffected in both anterior and posterior clones mutant for pka (d, f) or cos2 (g, i); note also that accumulation of Smo in anterior compartment cells near the A/P boundary is unaffected by the loss of Cos2 activity (j–l).
high levels of Smo protein accumulation relative to surrounding wild type cells (Fig. 3a – c). Since Ptc is a negative regulator of Hh signalling, it is possible that Smo levels might be stabilised by Hh signalling activity. In this case, activating Hh signalling by inactivating downstream negative regulators of the pathway ought to result in Smo stabilisation. Indeed it has been suggested that antagonising the activity of PKA, a negative regulator of the pathway, by over-expressing a dominant negative form of the protein, is sufficient to stabilise Smo in anterior compartment cells where it is usually downregulated (Alcedo et al., 2000). In contrast to this observation, we found that Smo levels are unaffected in clones of anterior cells mutant for PKA (Fig. 3d –f). It has also been suggested that Smo stability might be influenced by Cos-2 activity (Zhu et al., 2003; Hooper, 2003). However, we observed that clones of cells lacking Cos-2 activity showed no changes in the accumulation of the endogenous Smo protein (Fig. 3g – i). We note in particular that loss of Cos-2 activity led neither to an increase in Smo accumulation in Hh non-responding cells nor to a decrease in its accumulation in those cells that are responding to Hh at the A – P compartment boundary (Fig. 3j – l). Taken together, these data suggest that Smo protein is stabilised by Ptc inactivation rather than Hh signalling pathway activation and that Smo accumulation is a cause rather than a consequence of Hh signal transduction.
2.3. Activity of Smo gain of function mutations in imaginal discs Two oncogenic mutations in human Smo have been found in or close to the seventh transmembrane region (Xie et al., 1998) while several other constitutively active mutants have been selected on the basis of their resistance to the inhibitory effects of the alkaloid cyclopamine (Taipale et al., 2000). Experiments in cell culture have suggested that these proteins are ‘constitutively active’ because they escape Ptc repression (Xie et al., 1998). To investigate their effects in vivo, we introduced substitutions equivalent to either oncogenic mutant into the FLAG tagged smo cDNA (FLAG < SmoK580Q and FLAG < SmoW553L) and one of the cyclopamine resistant mutants into a cDNA encoding a GFP tagged variant of Smo (GFP < SmoA479Y) and used the GAL4-UAS system to misexpress each mutant cDNAs in the wing imaginal disc. By contrast to the wild type protein, the putative gain of function A479Y mutant form of GFP < Smo (GFP < SmoA479Y) accumulates at uniformly high levels in both compartments when expressed under ap-GAL4 control (Fig. 1b). Although expression of GFP < SmoA479Y by ap-GAL4 results in more extensive levels of ectopic dpp-lacZ upregulation (Fig. 1a,b), the resulting phenotype in the adult wing does not appear to be more severe than that
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
511
Fig. 4. The effects of Smo gain-of-function mutations in wing imaginal discs. Wings and wing imaginal discs in which FLAG-tagged Smo constructs have been misexpressed. (a) Ectopic expression of FLAG < Smo under the control of the 71B GAL4 driver results in ectopic veining around the second and third wing veins (L2 and L3), indicative of a mild activation of the Hh pathway. This can be seen more clearly in wing discs expressing FLAG < Smo under the control of 30A GAL4. The dpp-lacZ reporter gene is ectopically activated to a high level around the wing pouch (c), although no defect is observed in adult wings (b). Similar phenotypes are observed when oncogenic or constitutively active forms of Smo are misexpressed, although such cases exhibit a reduced level of ectopic dpp-lacZ expression compared with FLAG < Smo. (d–f) GFP < SmoA479Y; (g–i) FLAG < SmoK580Q; (j –l) FLAG < SmoW553L.
produced when GFP < Smo is expressed in the same way (data not shown). Misexpression of the FLAG tagged wild type protein driven by 30A causes ectopic activation of dpp-lacZ but has little if any effect on the differentiation of the wing; by contrast, misexpression using the 71B GAL4 (Brand and Perrimon, 1993) reproducibly results in the formation of extra veins in the anterior compartment of the wing blade (Fig. 4a). Surprisingly, we find that each of the putative gain of function mutant proteins, FLAG < SmoK580Q and FLAG < SmoW553L, behaves very similarly to the wild type protein when misexpressed in the same way; indeed, if anything they display a somewhat weaker activity, as judged both by the level of dpp-lacZ expression and the extent of extra venation formed when expressed under 71B control. 2.4. Smo function is disrupted by mutations in the amino-terminal and heptahelical domains Previous studies have identified several non-sense and two missense mutations in Smo; one of the latter causes a C–Y
substitution in the putative extra-cellular N-terminal domain (Chen and Struhl, 1998; Alcedo et al., 2000) whilst the other results in another C–Y substitution, in this case in the second transmembrane domain (Alcedo et al., 2000). To identify further critical residues in Smo, we undertook a screen for new EMS induced smo alleles (see Section 4). Amongst 8000 mutagenised second chromosomes screened, we identified seven new alleles on the basis of their lethality in trans to smo3 : Analysis of the cuticle patterns of homozygous embryos reveals a phenotypic series with smoIA3 ; smo2A and smo4DI exhibiting weaker phenotypes than smo3J ; smo1K ; smo4C2 and smo4F4 (Fig. 5). The first five of these alleles were characterised by single stranded conformation polymorphism (SSCP) analysis and/or DNA sequence analysis, the results of which are summarised schematically in Fig. 6. smo3J is associated with a stop codon at position 388 while smo1K has an 11 bp deletion after codon 437, resulting in a frameshift that cause premature termination after a further 15 codons. The three phenotypically weaker alleles by contrast, are all associated with missense mutations. In one of these, smo2A ; an Arg residue in the i3 loop, close to the boundary with TM6, is
512
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 5. Smo embryonic cuticle phenotypes. (a) Embryo homozygous for the smo2A allele showing a typical hypomorphic zygotic phenotype; note the partial fusion of abdominal denticle belts. (b) Embryo homozygous for the smo3 allele exhibiting a typical strong amorphic zygotic phenotype. (c) Partial rescue of the smo3 phenotype by prdGAL4 driven expression of FLAG < Smo; note the restoration of naked cuticle in alternate abdominal segments, corresponding to the domains of expression of the prdGAL4 driver.
substituted by a Cys. Notably, this Arg is conserved in all Frizzled and Smoothened proteins characterised to date. The smo4DI allele is associated with a Cys–Ser substitution at position 413 in the second putative extra-cellular loop, which is likely to disrupt the structure of the protein. The same substitution, this time in the Cys rich domain (CRD) of the putative extra-cellular N-terminal domain, is associated with the smoIA3 : While in vitro assays indicate that vertebrate Smo can retain activity in cultured mammalian cells in the absence of its CRD (Taipale et al., 2002), our finding that smoIA3 ; like smo2 (Alcedo et al., 2000) is associated with mutation of an Nterminal Cys residue, suggests that disulphide bridges within the CRD might play a critical role in effecting Smo activity in vivo. 2.5. The N-terminal extra-cellular domain is essential for Smo activity in vivo To investigate the requirements for the N-terminal extracellular domain (ECD) further, we generated DNFLAG <
Smo, a derivative of FLAG < Smo (Ingham et al., 2000) lacking all but 17 amino acids of the putative ECD. Ectopic expression of FLAG < Smo and of the DN derivative in wing imaginal discs using the GAL4 driver 30A results in uniformly high levels of FLAG immunoreactivity in both compartments, indicating that both transgenes are efficiently transcribed and translated throughout the enhancer trap domain (Fig. 7a,b). This contrasts with the post-translational modulation of endogenous Smo and of the GFP and HRP tagged forms described above. To confirm that the FLAG labelling accurately reflects the distribution of the transgene derived proteins, we also stained transgenic discs with an antibody raised against the C-terminus of Smo (Denef et al., 2000). This revealed an identical distribution in both cases (Fig. 7d, inset and data not shown), suggesting that the stability reflects the high levels of expression driven by the 30A GAL4 driver. Consistent with this, we found that GFP < Smo also accumulates at high levels in both compartments when expressed under 30A control (Fig. 7a, inset). Concomitant with the expression of FLAG, there is an increased accumulation of the full-length form of Ci and an ectopic activation of the dpp-lacZ reporter gene in anterior compartment cells expressing FLAG < Smo with 30A (Fig. 7b,c). Both of these effects are indicative of a derepression of the Hh pathway. By contrast, expression of DNFLAG < Smo in the same domain has no effect on Ci stability nor does it result in ectopic dpp-lacZ expression (Fig. 7d– f), confirming that the ECD is critical for Smo activity. 2.6. Progressive deletions reveal differential requirements for carboxy terminal regions of the Smo protein Since no mutations have been identified in the putative intracellular C-terminal domain (CTD) of Smo, we investigated the requirement for this region by deletion analysis. In the absence of any obvious motifs within CTD, we arbitrarily introduced stop codons at intervals of about 300 nucleotides, generating five different deletion constructs with progressively shorter intracellular tails. All carry the same N-terminal FLAG tag as the full-length FLAG < Smo. The largest of these, SmoDC0, ends immediately after the predicted seventh transmembrane region at residue T562 and thus lacks the entire CTD. Expression of this construct under 30A control results in high levels of FLAG accumulation but has no effect on dpplacZ expression (Fig. 8). SmoDC1 (DA638) similarly lacks any activity in this assay, indicating a requirement for the CTD that again contrasts with analysis of human Smo in tissue culture assays (Murone et al., 1999). By contrast, the three shorter deletion mutants, SmoDC2 (DT724), DC3 (DV837) and DC4 (DS939), all induce dpp-lacZ expression (Fig. 8). To test the activity of these constructs further, we performed rescue experiments on embryos mutant for both maternal and zygotic smo. Good restitution of ‘naked’
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
513
Fig. 6. Location of point mutations in the Smo protein. Schematic representation of the Smo protein showing its predicted heptahelical topology with long extra-cellular N-terminal, containing the CRD and intracellular CTD. Cys residues are indicated by green circles; mutated resides associated with different mutant alleles are indicated by red filled circles. The smoF5 and smo3 lesions were characterised by Chen and Struhl (1998) and the smo1 lesion by Alcedo et al. (2000); all other lesions shown were characterized in this study. Blue circles indicate the region of the C-terminal domain that is not conserved between the Drosophila and vertebrate forms of the protein. Smo1k not shown.
cuticle was seen with DC4, but already with DC3 rescue was variable and often incomplete (data not shown). Shorter constructs (three and two different lines were tested of DC1 and DC2, respectively) had no rescuing activity at all. Thus, this more stringent test revealed a requirement for the juxtamembrane half of the CTD, a region that has been highly conserved between Drosophila and vertebrates (see Quirk et al. (1997)).
3. Discussion The Smo protein is an essential component of the Hh signal transduction pathway; in all contexts analysed to date, inactivation of Smo renders cells incapable of responding to Hh signals (Lawrence et al., 1999b; Zhang and Kalderon, 2000; Zhang et al., 2001; Chen et al., 2001). Reciprocally, gain of function mutations have been isolated in human Smo that are sufficient to activate the Hh pathway
in the absence of ligand (Xie et al., 1998; Taipale et al., 2000; Hynes et al., 2000). Hh signalling regulates Smo activity via the multipass transmembrane receptor protein Ptc. In the absence of Hh ligand, Ptc represses Smo activity whereas when Hh binds to Ptc, Smo becomes active. Immunoprecipitation studies of vertebrate Smo from tissue culture cells co-expressing Ptc and Smo have suggested that this Ptc-dependent modulation of Smo activity is mediated by a direct interaction between the two proteins, a view consistent with the effects of high level Smo expression in cultured cells carrying a Shhresponsive reporter gene (Murone et al., 1999). Studies in Drosophila, however, have suggested that Ptc suppresses Smo activity by regulating the sub-cellular distribution and stability of the protein in a non-stoichiometric manner (Alcedo et al., 2000; Denef et al., 2000; Ingham et al., 2000; Taipale et al., 2002). The relatively weak effects of ectopic Smo expression in the wing imaginal disc described here and noted previously by other authors (Denef et al., 2000;
514
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Fig. 7. Effects of misexpression of FLAG < Smo and its N-terminally truncated derivative in wing imaginal discs. (a –c) Expression of FLAG < Smo under 30A control (a) results in ectopic accumulation of Ci155 (b) and ectopic activation of dpp-lacZ (c) around the edge of the presumptive wing blade in the anterior compartment, corresponding to those cells in which the transgene is expressed. (d–f) Similar 30A driven expression of DNFLAG < Smo (d) has no effect on Ci155 stability (e) or dpp-lacZ expression (f). (a, inset) Expression of GFP < Smo with 30A GAL4 shows uniform levels of GFP expression in both anterior and posterior compartments (compare with Fig. 6a) indicative of the strength of the 30A driver. (d, inset) Expression of DNFLAG:Smo detected with antibodies against the C-terminal domain of Smo. High levels of Smo expression in the anterior compartment demonstrate stable Smo expression from the transgene, even in the absence of the N-terminus, and are consistent with the strength of the 30A driver.
Martin et al., 2001; Zhu et al., 2003), presumably reflect this sub-stoichiometric regulation of the exogenously supplied protein by endogenous Ptc activity. Consistent with this, we find that ectopically expressed Smo protein is clearly subject to Ptc-mediated destabilisation, as revealed by the distribution of the GFP and YFP tagged forms of Smo. Nevertheless, when expressed at very high levels, as occurs for instance using the 30A GAL4 driver line, the Smo protein can escape the effects of Ptc activity and accumulate in anterior compartment cells where it activates Hh target gene activity. Even under these circumstances, however, the phenotypic effects of its over-expression are relatively mild. Smo shares limited sequence homology with members of the Frizzled family of Wnt receptors, in particular in the CRD of the putative N-terminal ECD that in Fz has been shown to be necessary and sufficient for Wnt binding (Bhanot et al., 1996). By contrast with Fz, there is no evidence that the ECD of Smo binds Wg (or other Wnts) or indeed Hh, consistent with the finding that Smo activity is independent of Hh in the absence of the inhibitory effects
of Ptc. Nevertheless, of the five Smo point mutants that we have identified, three affect residues in the N-terminal ECD, including two Cys residues that are highly conserved in all Fz family members, suggesting ECD and the CRD in particular plays some critical role in Smo function (see also Chen and Struhl (1998)). In line with this, we find that deletion of the ECD results in a loss of function of the protein when assayed both in our mutant rescue and imaginal disc ectopic expression systems. One explanation for this requirement could be that the N-terminus binds an as yet unidentified ligand that is required even in the absence of Ptc repressive activity, or indeed the secretion of which is inhibited by Ptc. Alternatively, the ECD may itself function as an activating ligand (see Chen and Struhl (1998)). The protease-activated G-protein coupled receptors (PARs) provide a precedent for such intramolecular receptor activation (Nystedt et al., 1994). PARs are activated by proteolytic cleavage of the ECD, the tip of the new amino terminus created by this cleavage acting as a tethered ligand by binding to the receptor core. We emphasise, however, that there is no evidence that Smo undergoes
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
515
Fig. 8. Progressive deletion of the C-terminal domain inactivates FLAG:Smo. The upper panels show the high level expression of transgenes encoding the C-terminal deletion derivatives of FLAG:Smo under 30A control, as revealed by anti-FLAG staining (green). Lower panels show the expression of dpp-lacZ (red) in discs expressing the respective transgenes.
a similar proteolytic activation. In contrast to our findings, Murone et al. (1999) have previously demonstrated that an N-terminally deleted form of human Smo retains activity when over expressed in 10T1/2 cells. The disparity between this result and our own findings may reflect differences in the levels of over-expression achieved in the in vivo versus the in vitro systems; it may be that at high enough levels, the mutant form of Smo can saturate the repressive activity of endogenous Ptc in the 10T1/2 cells, thus leading to the activation of the endogenous wild type Smo protein. Since no mutations have been identified in the CTD of Smo, we have also investigated the requirement for this region by deletion analysis. In contrast to the results of in vitro studies of human Smo suggesting that the terminal domain is dispensable (Murone et al., 1999) we found that deletion of the entire domain or even 80% of it completely inactivates the protein. Only deletions that preserve the juxtamembrane half of the C-terminal tail retain any activity; notably it is this part of the C-terminus that has been highly conserved between Drosophila and vertebrates (Quirk et al., 1997). Although these results demonstrate a requirement for carboxy terminal regions of the protein, they do not give any direct indications as to their molecular function. For example, structures in the carboxy terminus could be needed to keep other parts of Smo in the correct conformation. Alternatively, the carboxy terminus could bind directly to another protein or proteins that transduce
the Hh signal to the multimeric protein complex that regulates Ci stability and activation, or indeed to the multimeric complex itself. Amongst the new smo mutant alleles that we have isolated, we found one, smo2A ; to be a missense mutation that converts a highly conserved arginine to a cysteine at amino acid position 474 in the third intracellular loop. Residues at the C-terminal end of i3 in GPCRs have been shown to be critical for appropriate coupling to heterotrimeric G-proteins (Zhang et al., 2000) and loss of charged residues in this region of the angiotensin II receptor type I abolishes such coupling (Ohyama et al., 1992). While it has been shown that Frizzled proteins can transduce Wnt signals via heterotrimeric G-proteins (Malbon et al., 2001), the evidence for G-protein involvement in Smo activity remain equivocal (DeCamp et al., 2000; Murone et al., 1999; Norris et al., 2000). Both smo4DI and smo2A hypomorphic mutations reside in the heptahelical domain of the protein, the structural integrity of which has been implicated in Smo activity. Small molecule agonists and antagonists of vertebrate Smo have been shown to bind to this domain, implying its involvement in regulating the activity of the protein (Chen et al., 2002a,b). Exactly how these small molecules influence Smo activity is unclear, but one suggestion is that they may alter the conformation of the protein, respectively, towards a more active or inactive state (Chen et al., 2002b). In this view, the normal regulation of Smo by Ptc might involve the transport of a cellular small
516
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
molecule agonist/antagonist that similarly binds to Smo, altering its conformation. Alternatively, Ptc might act to redistribute Smo to a membrane compartment where it is susceptible to such an antagonist ligand or inaccessible to an agonist. Our analysis of the sub-cellular distribution of tagged forms of the Smo protein in Drosophila tissues is consistent with a role for Ptc in the intracellular trafficking of Smo (Martin et al., 2001; Strutt et al., 2001; Incardona et al., 2002). Specifically we find that Smo accumulates in the plasma membrane of cells that lack Ptc activity, whereas when co-expressed with Ptc, the same protein localises exclusively intracellularly, accumulating in distinct punctate structures. Similar results using the salivary gland as an assay system have recently been reported (Zhu et al., 2003). Significantly, our electron microscopy analysis of the HRP < Smo fusion proteins reveals high levels of HRP activity in the lysosomes of cells in which Ptc is active whereas in Ptc negative cells, the protein accumulates predominantly in the plasma membrane as well as in endosomes and unidentified presumptive transport vesicles. Analysis of the localisation and activity of putative hypermorphic forms of Smo might afford an insight into the mechanism of Smo function. Interestingly, when driven with ap-GAL4, the putative hypermorphic form of Smo, GFP < SmoA479, accumulates in the plasma membrane irrespective of Ptc activity and yields a stronger phenotype than GFP < Smo expressed under the same conditions, as assayed by dpp-lacZ expression. Surprisingly, however, the resulting adult phenotype appears to be no more severe in the case of the putative hypermorph than the wild type form. Furthermore, the over-expression of SmoA479Y as well as two other putative gain of function forms of Smo, SmoK580Q and SmoW553L, directed by the 30A and 71B GAL4 drivers resulted in lower levels of ectopic dpp-lacZ activation than those induced by wild type Smo. Consistent with this, the resulting adult phenotypes were also weaker in the case of the mutant forms of Smo than their wild type counterpart. We note, however, that Zhu et al. (2003) reported that the over-expression of certain putative hypermorphic forms of Smo results in stronger phenotypes than wild type Smo. The reason for these disparities is currently unclear. Finally, our observations that the cellular distribution of Smo is unaffected in imaginal disc clones lacking the activity of PKA or Cos2 (and hence constitutively transducing the Hh signal) has suggested Hh signalling is not sufficient to stabilise Smo and that Smo stabilisation is therefore more likely to be a cause, rather than an effect, of Hh signal transduction. However, Zhu et al. (2003) observed that over-expression of Cos2 in salivary gland cells can result in Smo destabilisation, suggesting that Cos2 activity is capable of influencing Smo levels. Taken together with our data, this suggests that antagonising Cos2 activity might be a necessary, but not sufficient, prerequisite for Smo stabilisation at the cell surface.
In conclusion, our data favour a mechanism whereby Ptc regulates Smo activity through inhibition of its accumulation in the plasma membrane, targeting it instead to the lysosomal pathway. In this view, Ptc might regulate Smo activity simply by modulating the levels of protein present in the cell. Alternatively, it may be that the sub-cellular location of Smo is critical for its activation, the plasma membrane perhaps providing an environment in which Smo is accessible to an intracellular agonist.
4. Materials and methods In vivo mutagenesis—cn bw sp males were mutagenised with ethylmethylsulfonylurea (EMS) using standard protocols (Ingham, 1984). Eight thousand chromosomes were screened for lethality in trans to the smo3 allele. The embryonic phenotypes of all non-complementing mutations isolated in this way were tested in trans to smo1 and smo2 alleles. In vitro mutagenesis—DNFlagsmo: A smo cDNA missing the sequence for the extra-cellular amino terminus was amplified using the sense primer 50 -TTATGCATTGC GATGCAAGGATCCACTC-30 , adding an Nsi I site, and 50 -TTCTCGAGCTATTTTGAAGGCAGCAATAACATT TTGAG-30 , adding an Xho I site after the translation stop codon. The fragment was cut with Nsi I and Xho I and cloned between the same sites in pSKII/SpFlag. FlagsmoDC constructs: A series of 30 deletion mutants of the full-length Flag-tagged smo cDNA (Ingham et al., 2000) were made using the following antisense primers (all containing a stop codon before an Xho I site): DT562: 50 -TTCTCGAGCTAAGTCTCAATTGAAGAA GGTG-30 ; DA638: 50 -TTCTCGAGCTATGCAGCCCAA GTTGATGA-30 ; DT742: 50 -TTCTCGAGCTAAGTACTG CTTTCCCTTCTCC-30 ; DV837: 50 -TTCTCGAGCTAT ACAACCACGTTAAGATCCTG-30 ; DS939: 5 0 -TTCT CGAGCTAGCTGTTAAGAAGGCTTCCTAC-30 , paired with suitable sense primers. Amplified fragments were digested with Xho I and Spe I or EcoR I, recognising internal sites, and cloned in pSKII/Flagsmo that had been cut with the same enzymes. UAS-Smo-EGFP construct: the stop codon of the smo cDNA was replaced by a KpnI site using standard PCR-based cloning. The resulting full-length smo cDNA with KpnI site at the C-terminus was cloned into pEGFP-N1 vector (Clontech). These constructs were introduced in w1118 host flies by P element-mediated transformation using standard methods. UAS-Boss-HRP-Smo construct: the signal sequences and HRP sequences were obtained from the Boss-HRP construct (Sunio et al., 1999) and 30 to the DNA encoding the signal sequence of Smo from glycine 37 was ligated using standard PCR-based cloning. The resulting sequence was M(1)…CH(33)agMQLT…NSGGklG(37)STTP. Boss signal sequences are in bold; the extra amino acids
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
generated by the fusion are indicated in lower case. The HRP sequences are italicised and bold letters indicate the amino acids from Smoothened. Smo gain of function mutants: point mutations resulting in amino acid substitutions were introduced following the manufacturer’s protocol. Smo cDNA was subcloned into a p-EGFP-N1 vector (clontech) in order to generate a GFP fusion protein with EGFP at the C-terminus (smo-EGFP). Smo cDNA was cut with the enzyme BclI to generate a fragment that was subcloned into bluescript pSKII. A point mutation (A479Y) was introduced by site directed mutagenesis (using Stratagene Quickchange kit). The mutated BclI fragment was then re-introduced into the smo-EGFP to make smoA479Y-EGFP. The smoA479YEGFP cDNA was released from the p-EGFP-N1 vector using NotI. All modified smo cDNAs were cloned into pUAST vector (Brand and Perrimon, 1993). The proofreading polymerase Pfu (Stratagene) was used throughout these manipulations and all new fragments were sequenced to check for any errors. Transgenic flies—A mixture of expression construct plasmid (< 450 ng/ml) and the transpose-encoding plasmid pUChsD2 –3 (< 150 ng/ml; plasmid kindly provided by Dr David I. Strutt) in water was injected into blastoderm yw embryos using a Leitz micromanipulator fitted to a Leica DM IL inverted microscope. Surviving adult flies were crossed to yw flies and transformant progeny identified by their pigmented eyes. Insertions were mapped to chromosome and balanced. Stocks were kept at 18 8C and crosses were done at 25 8C, unless otherwise indicated. Clonal analysis—First instar larvae were heat-shocked for 2 h at 38 8C on two consecutive days. Imaginal discs were dissected from wandering third instar larvae, fixed for 25 min in 4% paraformaldehyde in PBS and stained with antibodies to Smo (S. Cohen) Genotypes of the larvae were as follows: ptc clones: w hs-flp; FRT42D ptcS2/FRT42D ubi-GFP pka clones: w hs-flp; pka-c1 H2 FRT40A/ubi-GFP FRT40A cos2 clones: w hs-flp; FRT42D cos2k16101/FRT42D ubi-GFP Cuticle analysis—Unhatched first instar larvae were collected from apple juice plates, dechorionated in bleach and mounted directly in a 1:1 mixture of Hoyer’s medium and lactic acid. Wings from adult flies were dissected in isopropanol and mounted on microscope slides in Euparal. Antibodies—Monoclonal a-Flag antibodies M1 and M2 (Sigma) were used at a dilution of 1:200. M1 recognises an amino-terminal Flag tag sequence, providing a good indication that the transgenic protein has been correctly processed and inserted in the membrane. Monoclonal antiCi antibody 2A1 (kindly provided by Robert Holmgren) was used at 1:5 dilution. Monoclonal a-HA antibody 12CA5 (Boehringer) was used at 1:100 dilution. Rat polyclonal
517
anti-Smo antibody (kindly provided by Stephen Cohen) was used at 1:500. Rabbit a-b galactosidase antiserum was from Cappell and used at a dilution of 1:200. Fluorescent secondary antibodies were from Jackson Immunoresearch Laboratories Inc. or from Molecular Probes. Microscopy and image analysis—Images of fluorescently labelled wing discs were captured using a Leica TCS NT confocal microscope. The full depth of the discs was scanned in series of eight sections, which were then averaged using the ‘extended focus’ software function. Images were manipulated and the final figures composed using Adobe Photoshop CS software on Macintosh computers. 4.1. Transmission electron microscopy Embryos were fixed 10 min in heptane saturated with glutaraldehyde and devitellinised by hand in PBS with Tween (PTW). HRP activity was detected by incubation in DAB (0.5 mg/ml) with 0.003% H2O2 for 30 min. Embryos were washed with PBS and then processed for EM. Wing imaginal discs were dissected in PTW and incubated in DAB (0.5 mg/ml) with 0.003% H2O2 for 20 min. After washing in PTW, wing imaginal discs were fixed in 2% glutaraldehyde for 30 min and then washed in PBS and processed for EM. For EM, samples were embedded as described (Van Vactor et al., 1991). Ultrathin sections, approximately 70 – 90 nm thick, were post-stained with Reynold’s Lead Citrate and examined using a Philip CM10 Transmission Electron Microscope.
Acknowledgements We thank Bill Brown, Liz Jones, Debbie Sutton and Jeremy Quirk for help with the SSCP and sequence analysis and John Proctor for his expert assistance with the TEM. This work was supported by grants from Ontogeny inc. and the Wellcome Trust to PWI. AAS is funded by an MRC postgraduate studentship. The Centre for Developmental Genetics confocal microscopy facility is funded by Yorkshire Cancer Research.
References Alcedo, J., Ayzenzon, M., Vonohlen, T., Noll, M., Hooper, J., 1996. The Drosophila Smoothened gene encodes a 7-pass membrane-protein, a putative receptor for the Hedgehog signal. Cell 86, 221–232. Alcedo, J., Zou, Y., Noll, M., 2000. Post-transcriptional regulation of Smoothened is part of a self-correcting mechanism in the Hedgehog signaling system. Mol. Cell 6, 457 –465. Bhanot, P., Brink, M., Samos, C.H., Hsieh, J.-C., Wang, Y., Macke, J.P., et al., 1996. A new member of the Frizzled family from Drosophila functions as a Wingless receptor. Nature 382, 225–230. Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118, 401–415.
518
Y. Nakano et al. / Mechanisms of Development 121 (2004) 507–518
Chen, Y., Struhl, G., 1998. In vivo evidence that Patched and Smoothened constitute distinct binding and transducing components of a Hedgehog receptor complex. Development 125, 4943–4948. Chen, W., Burgess, S., Hopkins, N., 2001. Analysis of the zebrafish Smoothened mutant reveals conserved and divergent functions of Hedgehog activity. Development 128, 2385–2396. Chen, J., Taipale, J., Cooper, M., Beachy, P., 2002a. Inhibition of Hedgehog signaling by direct binding of cyclopamine to Smoothened. Genes Dev. 16, 2743– 2748. Chen, J., Taipale, J., Young, K., Maiti, T., Beachy, P., 2002b. Small molecule modulation of Smoothened activity. Proc. Natl. Acad. Sci. USA 99, 14071–14076. DeCamp, D., Thompson, T., de Sauvage, F., Lerner, M., 2000. Smoothened activates Ga i-mediated signaling in frog melanophores. J. Biol. Chem. 275, 26322–26327. Denef, N., Neubuser, D., Perez, L., Cohen, S., 2000. Hedgehog induces opposite changes in turnover and subcellular localization of Patched and Smoothened. Cell 102, 521–531. van den Heuvel, M., Ingham, P.W., 1996. Smoothened encodes a serpentine protein required for Hedgehog signaling. Nature 382, 547 –551. Hooper, J.E., 2003. Smoothened translates Hedgehog levels into distinct responses. Development 130, 3951–3963. Hynes, M., Ye, W., Wang, K., Stone, D., Murone, M., Sauvage, F., Rosenthal, A., 2000. The seven-transmembrane receptor Smoothened cell-autonomously induces multiple ventral cell types. Nat. Neurosci. 3, 41 –46. Incardona, J., Gruenberg, J., Roelink, H., 2002. Sonic Hedgehog induces the segregation of Patched and Smoothened in endosomes. Curr. Biol. 12, 983 –995. Ingham, P.W., 1984. A gene that regulates the Bithorax complex differentially in larval and adult cells of Drosophila. Cell 37, 815–823. Ingham, P.W., McMahon, A.P., 2001. Hedgehog signaling in animal development: paradigms and principles. Genes Dev. 15, 3059–3087. Ingham, P., Nystedt, S., Nakano, Y., Brown, W., Stark, D., van den Heuvel, M., Taylor, A., 2000. Patched represses the Hedgehog signalling pathway by promoting modification of the Smoothened protein. Curr. Biol. 10, 1315–1318. Lawrence, P., Casal, J., Struhl, G., 1999b. The Hedgehog morphogen and gradients of cell affinity in the abdomen of Drosophila. Development 126, 2441–2449. Malbon, C., Wang, H., Moon, R., 2001. Wnt signaling and heterotrimeric G-proteins: strange bedfellows or a classic romance? Biochem. Biophys. Res. Commun. 287, 589–593. Martin, V., Carrillo, G., Torroja, C., Guerrero, I., 2001. The sterol-sensing domain of Patched protein seems to control Smoothened activity through Patched vesicular trafficking. Curr. Biol. 11, 601 –607. McMahon, A.P., Ingham, P.W., Tabin, C., 2003. The developmental roles and clinical significance of Hedeghog signalling. Curr. Topics Dev. Biol. 53, 1–114. Murone, M., Rosenthal, A., de Sauvage, F., 1999. Sonic Hedgehog signaling by the Patched–Smoothened receptor complex. Curr. Biol. 9, 76–84.
Norris, W., Neyt, C., Ingham, P., Currie, P., 2000. Slow muscle induction by Hedgehog signalling in vitro. J. Cell Sci. 113, 2695–2703. Nystedt, S., Emilsson, K., Wahlestedt, C., Sundelin, J., 1994. Molecular cloning of a potential proteinase activated receptor. Proc. Natl. Acad. Sci. USA 91, 9208–9212. Ohyama, K., Yamano, Y., Chaki, S., Kondo, T., Inagami, T., 1992. Domains for G-protein coupling in angiotensin II receptor type I: studies by site-directed mutagenesis. Biochem. Biophys. Res. Commun. 189, 677–683. Quirk, J., van den Heuvel, M., Henrique, D., Marigo, V., Jones, T., Tabin, C., Ingham, P., 1997. The Smoothened gene and Hedgehog signal transduction in Drosophila and vertebrate development. Cold Spring Harbor Symp. 62, 217 –226. Strutt, H., Thomas, C., Nakano, Y., Stark, D., Neave, B., Taylor, A., Ingham, P., 2001. Mutations in the sterol sensing domain of Patched suggest a role for vesicular trafficking in Smoothened regulation. Curr. Biol. 11, 608 –613. Sunio, A., Metcalf, A., Kra¨mer, H., 1999. Genetic dissection of endocytic trafficking in Drosophila using a horseradish peroxidase-bride of sevenless chimera: hook is required for normal maturation of multivesicular endosomes. Mol. Biol. Cell 10, 847 –859. Taipale, J., Chen, J., Cooper, M., Wang, B., Mann, R., Milenkovic, L., et al., 2000. Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature 406, 1005–1009. Taipale, J., Cooper, M.K., Maiti, T., Beachy, P.A., 2002. Patched acts catalytically to suppress the activity of Smoothened. Nature 418, 892 –897. Van Vactor, D.L. Jr., Cagan, R., Kra¨mer, H., Zipursky, S., 1991. Induction in the developing compound eye of Drosophila: multiple mechanisms restrict R7 induction to a single retinal precursor cell. Cell 67, 1145–1155. Varga, Z.M., Amores, A., Lewis, K.E., Yan, Y.L., Postlethwait, J.H., Eisen, J.S., Westerfield, M., 2001. Zebrafish Smoothened functions in ventral neural tube specification and axon tract formation. Development 128, 3497–3509. Xie, J., Murone, M., Luoh, S., Ryan, A., Gu, Q., Zhang, C., et al., 1998. Activating Smoothened mutations in sporadic basal-cell carcinoma. Nature 391, 90–92. Zhang, Y., Kalderon, D., 2000. Regulation of cell proliferation and patterning in Drosophila oogenesis by Hedgehog signaling. Development 127, 2165–2176. Zhang, M., Zhao, X., Chen, H., Catt, K., Hunyady, L., 2000. Activation of the AT1 angiotensin receptor is dependent on adjacent apolar residues in the carboxyl terminus of the third cytoplasmic loop. J. Biol. Chem. 275, 15782–15788. Zhang, X., Ramalho-Santos, M., McMahon, A., 2001. Smoothened mutants reveal redundant roles for Shh and Ihh signaling including regulation of L/R asymmetry by the mouse node. Cell 105, 781 –792. Zhu, A., Zheng, L., Suyama, K., Scott, M., 2003. Altered localization of Drosophila Smoothened protein activates Hedgehog signal transduction. Genes Dev. 17, 1240–1252.