Abstract
Contrary to the canonical model of Hh signaling, we find that cells genetically lacking Ptch1 and Ptch2 remain responsive to ShhN. These cells retain the ability to migrate towards a source of ShhN, while expression of ShhN results in a robust activation of the transcriptional Hh response; both occur in a Smo-dependent manner. This activation of Hh responses does not require binding to the co-receptors Boc, Cdo, or Gas1, nor Shh autocatalytic processing, as the cholesterol moiety on Shh impedes signaling. Shh mutants unable to bind their cognate receptors, or fail to undergo proper processing, nevertheless retain their ability to activate the Hh response both in vivo and in vitro. Together, our findings support a model in which the role of Ptch1/2 as an inhibitor of Smo intersects with a Shh-mediated Smo activation event that is independent of known Shh receptors.
Introduction
Sonic Hedgehog (Shh) is a signaling molecule indispensable for vertebrate embryonic development and adult stem cell niche maintenance (Ingham and McMahon, 2001). Impaired regulation of the Hedgehog (Hh) pathway is a known cause of various birth defects and diseases (Bale, 2002).
In the absence of the Hh ligands, the receptors Patched1 (Ptch1) and Patched2 (Ptch2) inhibit the signal transducer Smoothened (Smo) through a sub-stoichiometric mechanism (Alfaro et al., 2014; Taipale et al., 2002; Zhang et al., 2001a). Ptch1/2 belong to the Resistance, Nodulation, and Division (RND) family of transmembrane transporters, and their inhibition requires the putative proton anti-porter activity that characterizes these molecules (Taipale et al., 2002). In general, RNDs mediate the secretion of small lipophilic and amphipathic molecules (Nikaido and Zgurskaya, 2001; Nikaido and Takatsuka, 2009), and it has been proposed that Ptch1/2 re-localize a sterol that modifies Smo activity (Huang et al., 2016; Taipale et al., 2002).
The binding of extracellular Shh requires Ptch1/2 in conjunction with the co-receptors, Boc, Cdo, and Gas1 (Allen et al., 2011; Izzi et al., 2011; Marigo et al., 1996). Shh is internalized during signaling, causing a change in the subcellular distribution and activity of Smo (Aanstad et al., 2009; Bijlsma et al., 2012; Incardona et al., 2002; 2000; Milenkovic et al., 2009). How these events regulate Smo activity remains unclear. The Hh pathway is upregulated in Ptch1 null mice and neuralized Ptch1LacZ/LacZ;Ptch2-/- mouse embryonic stem cells, supporting the canonical model that Smo activity is principally regulated by changes in Ptch1/2-mediated inhibition (Alfaro et al., 2014; Goodrich et al., 1997). A tenet of this model is that Smo is constitutively active in the absence of Ptch1/2 function.
Smo is a Class F G-protein coupled receptor (GPCR) and belongs to the superfamily of receptors predominantly defined by Frizzleds (Frz), the canonical receptors of the Wnt signaling pathway (Bhanot et al., 1996; Kristiansen, 2004). Smo and Frzs share over 25% sequence identity and harbor a conserved extracellular, amino-terminal Cysteine Rich Domain (CRD) and seven hydrophobic domains. The CRD of Frz binds to Wnt through two distinct binding sites, one of which is a protein-lipid interface, to initiate signal transduction (Janda et al., 2012). Similarly, the CRD of Smo has been shown to bind to a variety of sterols (Myers et al., 2013; Nachtergaele et al., 2013; Nedelcu et al., 2013), including cholesterol (Huang et al., 2016), which can modulate Smo activity. Smo can also be regulated by small molecule agonists and antagonists that target its membrane-exposed heptahelical domain. This heptahelical domain of Smo is thought to be the target of the inhibitory effect of Ptch1 (Chen et al., 2002a; 2002b; Wang and McMahon, 2013).
Our results support earlier observations that Smo activity can be regulated in a Ptch1/2 independent manner, by demonstrating that ShhN can mediate the activation of Smo in cells lacking Ptch1/2. Ptch1 and Ptch2 are dispensable for a cell’s ability to migrate towards a source of ShhN, a Hh response that nevertheless requires Smo. Although exogenously supplied ShhN does not induce a transcriptional Hh response in cells lacking Ptch1/2, transfection of ShhN causes a significant Smo-dependent activation of the transcriptional response. The mechanism regulating cell-autonomous Smo activation is fundamentally different from non-cell autonomous activation via the canonical Shh receptors. Shh mutants unable to bind to canonical Hh receptors, or unable to undergo post-translational processing, retain their ability to induce the Hh response after expression in cells lacking Ptch1/2, as well as after expression in the developing neural tube. Our work reveals a more complex model of Shh signal transduction in which the role of Ptch1 as a catalytic inhibitor of Smo activity intersects with a relatively direct interaction between Shh and Smo, together regulating the definitive level of pathway activation.
Results
The CRD of Smo is not required for its inhibition by Ptch1/2
Sterols have long been proposed to be the cargo which Ptch1 transports across membranes to regulate Smo activity, leaving open the possibility that the CRD of Smo is the target of Ptch1 function. While this issue remains unresolved, it has been shown that a form of Smo lacking its CRD (SmoΔCRD) has reduced sensitivity to Shh signaling (Myers et al., 2013; Nachtergaele et al., 2013) and increased constitutive activity (Byrne et al., 2016). To determine whether Ptch1 targets the CRD of Smo, we used Ptch1LacZ/LacZ;Ptch2-/- fibroblasts derived from mouse embryonic stem cells (mESCs) (Roberts et al., 2016) to assess Ptch1 function in the absence of endogenous Ptch1/2 activity. Transfection of either Smo or SmoΔCRD into Ptch1LacZ/LacZ;Ptch2-/- cells moderately raised the level of Hh pathway activity (Figure 1A), measured using a well-characterized Gli-Luciferase reporter (Taipale et al., 2002). Co-transfection of Smo or SmoΔCRD together with Ptch1 or Ptch1ΔL2 (a dominant inhibitory form of Ptch1 (Briscoe et al., 2001)), predictably lowered Hh pathway activity caused by Smo expression. However, co-transfection of Smo or SmoΔCRD with forms of Ptch1 that contain mutations within the putative antiporter domain, Ptch1D499A and Ptch1ΔL2-D499A, were less effective at inhibiting the Hh pathway activity caused by Smo overexpression (Figure 1A). Thus Smo, including forms lacking its CRD, remains subject to the inhibitory effects of the proton antiporter activity of Ptch1, demonstrating that the CRD of Smo is not the target of Ptch1-mediated inhibition.
To verify our observation in vivo, we assessed the ability of Ptch1 to inhibit SmoΔCRD in the developing embryo. Electroporation of the chick neural tube is a well-established procedure to introduce transgenes into a normally developing embryo (Itasaki et al., 1999). Electroporation of Smo into stage 10-11 (Hamburger and Hamilton, 1951) chick neural tubes did not affect the Hh response pathway (Figure 1B-D). However, electroporation of SmoΔCRD resulted in an expansion of the ventral neural markers Nkx2.2 and Mnr2 (Figure1E-G), indicative of an ectopic activation of the Hh response pathway. Intrinsic activation of SmoΔCRD has been observed before (Byrne et al., 2016), which indicates that the CRD of Smo is able to regulate its downstream activity. We find that the activity of SmoΔCRD remains subject to Ptch1-mediated inhibition in vivo as well, as electroporation of SmoΔCRD together with Ptch1ΔL2, reversed the phenotypic effects of SmoΔCRD expression (Figure 1H-J). SmoΔCRD has been shown to constitutively localize to the primary cilium, the cellular compartment where it mediates the transcriptional Hh response (Aanstad et al., 2009). The entry of Smo into the cilium is required for transcriptional pathway activation, and we find that electroporation of a form of SmoΔCRD that cannot localize to the primary cilium, SmoΔCRD-CLD (Aanstad et al., 2009; Corbit et al., 2005), did not result in ectopic activation of the pathway in vivo (Figure 1K-M). These results demonstrate that the CRD of Smo is not required for its inhibition by Ptch1, and thus lends support to the idea that the heptahelical domain of Smo is the target for Ptch1 inhibition.
SmoΔCRD is intrinsically active in vivo (Figure 1E-G), while the activity of Smo and SmoΔCRD in Ptch1LacZ/LacZ;Ptch2-/- cells is indistinguishable (Figure 1A). To evaluate Smo activity in cells lacking all Ptch activity, we derived fibroblasts from Ptch1LacZ/LacZ;Ptch2-/-;Smo-/- mESCs (Robertson et al., 2016). We assessed the intrinsic activity of Smo and its mutant derivatives in these cells by transfecting the Gli-Luciferase reporter with Smo, SmoΔCRD, SmoΔCRD-CLD, SmoM2 (an intrinsically active mutant (Xie et al., 1998)), SmoL112D-W113Y (a mutant unable to bind sterols within the CRD (Nachtergaele et al., 2013)), or Drosophila Smo. Remarkably, none of these mutant forms of Smo induced the Hh response better than wild type (Figure 1N). Overall, SmoΔCRD-CLD, SmoL112D-W113Y and Drosophila Smo were less able to induce the transcriptional Hh response compared to Smo, SmoΔCRD, and SmoM2 (Figure 1N). This observation is consistent with the notion that SmoM2 is refractory to Ptch1/2 regulation, and suggests that an additional activation event may be required to further induce a Hh response.
We next tested whether we could further modulate Smo activity with SAG, an agonist that targets the heptahelical domain, in the absence of Ptch1/2 activity. We measured the Hh response using both the Gli-Luciferase reporter and the endogenous Ptch1:LacZ reporter contained within Ptch1+/LacZ and Ptch1LacZ/LacZ;Ptch2-/- fibroblasts. We found that both transcriptional reporters revealed identical results; Ptch1LacZ/LacZ;Ptch2-/- cells were completely insensitive to SAG, a powerful activator of Smo in cells that retain Ptch1/2 activity (Figure 1O). These results indicate that Ptch1/2 function is required for SAG-mediated activation of Smo; and thus might act as an antagonist to the inhibitory activity of Ptch1.
ShhN can activate the Hh pathway independent of Ptch1/2 function
As a central ligand in the activation of the Hh response, we assessed whether Shh can affect Smo activity in the absence of Ptch1/2. We have previously demonstrated that Shh-mediated chemotaxis involves the activation of Smo that is quick (in the order of minutes), independent of the primary cilium (Bijlsma et al., 2012), and independent of Ptch1 (Alfaro et al., 2014); leaving open the possibility that Ptch2 serves as the primary receptor. We tested the ability of Ptch1LacZ/LacZ;Ptch2-/- fibroblasts to migrate towards a localized source of ShhN (a soluble form of Shh) using a modified Boyden chamber (Chen, 2005). We found that Ptch1LacZ/LacZ;Ptch2-/- cells were indistinguishable from Ptch1LacZ/LacZ cells in their ability to migrate towards ShhN (Figure 2A). Shh chemotaxis was largely abolished in Ptch1LacZ/LacZ;Ptch2-/-;Smo-/- cells, although these cells were retained the ability to migrate towards FCS, our positive control (Figure 2A). These results demonstrate that all Ptch1/2 function is dispensable for the activation of Smo by exogenous ShhN. As the Shh co-receptor Boc has been implicated in the Shh-mediated guidance of commissural growth cones towards the floor plate (Okada et al., 2006), we made a fibroblast line that, besides Ptch1/2, also lack Boc, its paralog Cdo, as well as the co-receptor Gas1. We found that Ptch1LacZ/LacZ;Ptch2-/-;Shh-/-;Boc-/-;Cdo-/-;Gas1-/- cells, like all cell lines tested, retain their ability to migrate towards FCS; however, their ability to migrate towards ShhN is impaired, but not completely abolished (Figure 2A).
These findings are not readily reconciled with our earlier observation that Ptch1LacZ/LacZ;Ptch2-/- cells cannot upregulate the transcriptional Hh response when exposed to exogenous ShhN, unless transfected with Ptch1 (Roberts et al., 2016). Since Ptch1 mediates the internalization of Shh during signaling (Incardona et al., 2000), we tested whether we could circumvent this function of Ptch1 by providing ShhN intracellularly through transfection of a ShhN construct. Using a panel of cell lines lacking various combinations of Ptch1 and Ptch2, we found that transfection of ShhN invariably induced an upregulation of the transcriptional Hh response regardless of the presence or absence of Ptch1/2 (Figure 2B, C). The induction of the Hh response was measured using both a transfected Gli-Luciferase reporter (Figure 2B), as well as the genetically encoded Ptch1:LacZ reporter (Goodrich et al., 1997) (Figure 2C). The finding that Ptch1LacZ/LacZ;Ptch2-/- cells cannot activate the transcriptional Hh pathway in response to extracellular ShhN indicates that signaling is occurring intracellularly in cells lacking Ptch1/2, and is not the result of non-cell autonomous signaling; henceforth we will refer to this as “cell-autonomous activation”. In all cases, we found that the Smo specific inhibitor Vismodegib (Rudin, 2012) significantly blocked the induction, demonstrating that the transcriptional Hh response caused by ShhN transfection is mediated by Smo (Figure 2B, C). We found a high level of cilial occupation by Smo:GFP (Figure 2D) in Ptch1LacZ/LacZ;Ptch2-/- cells, consistent with earlier observations in Ptch1-/- cells (Rohatgi et al., 2007). Transfection of ShhN in Ptch1LacZ/LacZ;Ptch2-/- cells caused a small, but significant increase in the cilial occupation of Smo::GFP (Figure 2D), a demonstrated output of the transcriptional Hh response. Together these results demonstrate that ShhN can mediate activation of the Hh response pathway via a mechanism that does not require Ptch1/2 function.
We found that the Ptch1LacZ/LacZ fibroblast cell line has a higher level of intrinsic pathway activity compared to the Ptch1LacZ/LacZ;Ptch2-/- cell line, as measured by both the Gli-luciferase and Ptch1:LacZ reporters (Figure 2, Figure Supplement 1). Our results that Vismodegib treatment lowers the basal level of pathway activity in Ptch1LacZ/LacZ cells is consistent with this observation. However, the induction of transcriptional pathway activity caused by ShhN transfection in Ptch1LacZ/LacZ cells, as well as the ability of these cells to migrate towards a source ShhN, demonstrate that the absence of Ptch1 does not result in full activation of the Hh response; necessitating additional mechanisms by which Smo can be activated in the absence of Ptch1. The Ptch1LacZ/LacZ fibroblast line was derived from mutant embryos, while we derived the compound mutant fibroblast line from embryonic stem cells. The methods by which these cells were derived may explain the differences in their basal level of Hh pathway activity; epigenetically, these cells are expected to be quite different. This notion is supported by the characteristics of a Ptch1 null, Ptch2 containing fibroblast line (Ptch1-/-;Disp1-/-;Shh-/-) we derived from mESCs; which, unlike the Ptch1LacZ/LacZ line, retains a low level of intrinsic pathway activity similar to the Ptch1LacZ/LacZ;Ptch2-/- cell line, however remains equally sensitive to ShhN transfection (Figure 2B).
Despite differences in the levels of endogenous Hh pathway activity, we found that co-transfection of Gli-Luc with Ptch1 or Ptch1ΔL2 effectively inhibited the transcriptional Hh pathway in both Ptch1LacZ/LacZ and Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 2, Figure Supplement 2A). Forms of Ptch1 lacking the putative antiporter function, Ptch1D499A and Ptch1ΔL2-D499A, were less effective inhibitors of the Hh response pathway than their antiporter-competent counterparts. Transfection of Gli3PHS, a dominant inhibitory form of Gli3 (Meyer and Roelink, 2003), efficiently repressed the level of Hh pathway activity in both cells lines (Figure 2, Figure Supplement 2A), while overexpression of Gli1 caused a significant 10-fold induction (data not shown). These results demonstrate that, regardless of the basal state of the pathway, both cells lines predictably respond to inhibitory elements and thus retain all of the proper components of Hh pathway regulation. To test whether the transcriptional Hh response caused by ShhN transfection is subject to the same regulation, we co-transfected ShhN with either Ptch1ΔL2 or Gli3PHS. We found that the induction caused by ShhN transfection was inhibited by Ptch1ΔL2, demonstrating that ShhN-activated Smo remains subject to Ptch1 inhibition in both Ptch1LacZ/LacZ and Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 2, Figure Supplement 2B). The relative decrease of Smo activity caused by Ptch1 inhibition is similar under all conditions, which supports the idea that Ptch1 is a non-competitive inhibitor of Smo activation by ShhN. Similarly, Gli3PHS was able to repress the induction caused by ShhN transfection, implicating the canonical Gli transcription factors as the mediators of this Hh response. These results validate the Gli-Luciferase reporter as a faithful representation of the state of the Hh pathway in these cells.
Ptch1/2-independent cell-autonomous activation of the Hh response does not require Boc, Cdo or Gas1
Boc, Cdo, and Gas1 have been shown to function in an overlapping manner as obligate Shh co-receptors with Ptch1/2 during non-cell autonomous Hh signaling (Allen et al., 2011; Izzi et al., 2011; Tenzen et al., 2006). The observation that Ptch1 and Ptch2 are dispensable for the activation of Smo after ShhN transfection raises the question whether the Shh co-receptors are involved in detecting ShhN in cells independently of Ptch1/2 during cell-autonomous signaling. We addressed the requirement for Boc, Cdo and Gas1 in two complementary ways. 1) We assessed whether Shh mutants unable to bind to the co-receptors could activate the Hh response cell-autonomously, and 2) we tested whether cells lacking the co-receptors remained sensitive to ShhN transfection.
ShhN-E90A, a previously characterized ShhN mutant, cannot bind to either Boc, Cdo, or Gas1, and is unable to activate the Hh response when applied to wildtype cells (Izzi et al., 2011). To independently assess the requirement of Ptch1/2 binding in cell-autonomous pathway activation, we utilized a ShhN mutant with a disrupted zinc coordination site predicted to perturb Ptch1 binding, ShhN-H183A (Fuse et al., 1999; Goetz et al., 2006). We first tested the ability of ShhN-E90A and ShhN-H183A to signal non-cell autonomously (in trans), by transfecting Ptch1 and Gli-Luc into Ptch1LacZ/LacZ;Ptch2-/- “reporter” cells and co-culturing them with Ptch1LacZ/LacZ;Ptch2-/- cells independently transfected with ShhN, ShhN-E90A, or ShhN-H183A (Figure 3A, seafoam green). We confirmed our earlier observations that Ptch1-expressing Ptch1LacZ/LacZ;Ptch2-/- reporter cells activated the Hh response when co-cultured with ShhN expressing cells (Figure 3C). Reporter cells did not activate the Hh response when cultured with ShhN-E90A expressing cells (Figure 3C), corroborating previously published data that ShhN-E90A is unable to mediate Hh signaling in trans (Izzi et al., 2011). Consistent with its predicted inability to bind to Ptch1, we find that ShhN-H183A did not activate the Hh response in reporter cells (Figure 3C). These results support previous observations that activation of the transcriptional response by ShhN in the extracellular space (in trans signaling) requires the binding of Shh to Ptch1/2 in conjunction with the co-receptors.
ShhN is a widely used mutant form of Shh, it is freely soluble while retaining its inductive properties (Bumcrot et al., 1995a; Roelink et al., 1995). A significant distinction between ShhN and fully matured Shh is that Shh has a cholesterol moiety that serves as a membrane anchor (Yang et al., 1997). The distinct differences in distribution and extracellular transport between cholesterol-modified and soluble forms of Shh might have significant ramifications for their activities. This led us to test the ability of Shh and mutant derivatives expected to be cholesterol-modified (ShhE90A and ShhH183A) to signal in trans. Ptch1LacZ/LacZ;Ptch2-/- reporter cells (transfected with Gli-Luc and Ptch1) activated the Hh response pathway when co-cultured with Shh and ShhE90A expressing cells, but not with ShhH183A expressing cells (Figure 3E). This result indicates that the cholesterol modification of Shh is able to compensate for the inability to bind Cdo, Boc or Gas1 when signaling in trans. The inability of ShhH183A to signal in trans may be due, in part, to its poor autocatalytic cleavage, which is directly linked to its modification with a cholesterol moiety (Figure 3B) (Goetz et al., 2006).
We assessed the role of co-receptor binding during cell-autonomous Hh signaling by transfecting mutant forms of ShhN, ShhN-E90A and ShhN-H183, into Ptch1LacZ/LacZ;Ptch2-/- cells together with the Gli-Luc reporter (Figure 3A, lilac). We found that, similar ShhN, both ShhN-E90A and ShhN-H183A significantly activated the Hh pathway in Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 3D); indicating that neither Boc, Cdo, nor Gas1 are required for cell-autonomous pathway activation. We next evaluated whether the cholesterol modification on Shh mutants affects cell-autonomous pathway activation by transfecting Shh constructs together with the Gli-Luc reporter in Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 3A, lilac). We found that transfection of Shh into Ptch1LacZ/LacZ;Ptch2-/- cells results in a minute increase of pathway activity (Figure 3F), suggesting that the cholesterol modification on Shh is a major impediment to cell-autonomous activation of the Hh response. In contrast, transfection of ShhE90A or ShhH183A resulted in a significantly greater pathway induction (Figure 3F). The poor autoproteolysis observed with ShhH183A, which would prevent its cholesterol modification, makes this result difficult to interpret; however, all together these results suggest that Shh interaction with Boc, Cdo or Gas1 impedes cell-autonomous Hh signaling with cholesterol-modified Shh. We confirmed that transfection of these Shh mutants activates a cell-autonomous Hh response in Ptch1LacZ/LacZ in a manner similar to Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 3, Figure Supplement 1); again we found that wildtype Shh is a poor inducer of cell-autonomous pathway activation compared to its mutant counterparts.
To further verify the observation that Boc, Cdo and Gas1 are not a required component of the cell-autonomous Hh response, we generated a Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- mESC line and a derived fibroblast line of the same genotype. Upon differentiation into neuralized embryoid bodies, these cells behave similarly to their parental Ptch1LacZ/LacZ;Ptch2-/-;Shh-/- line (Roberts et al., 2016), and express ventral neural markers indicative of an activated Hh response (Figure 3G). The neural phenotype of differentiated Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- mESCs demonstrates that pathway activation in these cells does not require expression of Shh or the Shh (co-)receptors. Transfection of ShhN or any other Shh mutant into Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- fibroblasts cells induced a significant increase of the Hh response (Figure 3H). The independently derived Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- and Ptch1LacZ/LacZ;Ptch2-/- fibroblasts behave nearly indistinguishably in this assay.
To assess whether cell-autonomous pathway activation is a phenomenon that occurs across all Hh paralogs, we tested the ability of Indian Hedgehog (Ihh) and Desert Hedgehog (Dhh) to activate the Hh pathway after transfection. The soluble forms IhhN and DhhN are similar to ShhN in their ability to induce the Hh response cell-autonomously, whereas wildtype Ihh and Dhh are significantly more potent than Shh (Figure 3I). These findings demonstrate that cell-autonomous pathway activation is a characteristic of all vertebrate Hedgehog paralogs.
Together our data confirms that the extracellular receptors Boc, Cdo, and Gas1 are not required for pathway activation mediated by ShhN expression in Ptch1/2 null cells. As we found before in Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 3F), expression of wild type Shh in Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- only marginally induces the Hh pathway, a further demonstration that the cholesterol moiety prevents cell autonomous signaling by Shh (Figure 3I). This result demonstrates that it is not Boc/Cdo/Gas1 binding that impedes cell-autonomous activation by cholesterol-modified Shh; further analysis is required to understand the activity of ShhE90A
Shh mutants unable to bind to canonical receptors induce the Hh response in the developing neural tube
Both Ptch1 and Ptch2, as well as Boc, Cdo, and Gas1 have been shown to play central roles in regulation of the Hh responses in vivo (Allen et al., 2011; Goodrich et al., 1997; Izzi et al., 2011). We assessed whether expression of Shh mutants can activate the Hh response independent of (co)-receptor interaction in the developing neural tube of chicken embryos. We electroporated stage 10-11 (Hamburger and Hamilton, 1951) chick neural tubes with ShhN, ShhN-E90A, or ShhN-H183 together with GFP. We found that all ShhN mutants are able to activate the Hh pathway as assessed by changes in Shh-mediated dorsal-ventral patterning (Figure 4A-J). Expression of ShhN, ShhN-E90A, and ShhN-H183A causes extensive dorsal expansion of the Nkx2.2 and Mnr2 domains (Figure 4A-F), as well as repression of Pax7 (Figure 4G-J). These results support our finding that binding to the canonical (co)-receptors is not a necessary component for activation of the Hh response in vivo.
In order to assess more biologically relevant forms of Shh, we electroporated Shh, ShhE90A, or ShhH183A together with GFP in developing chick neural tubes. Expression of Shh or ShhE90A caused an expansion of Nkx2.2 and Mnr2 domains, as well as a repression of Pax7 (Figure 5A, B, E, F, I, J), indicating that these forms of Shh are active. However, electroporation of ShhH183A does not result in extensive activation of the pathway (Figure 5C, G, K). It is unclear why ShhH183A is able to activate the pathway in vitro but not in vivo. One possibility is that the inability of ShhH183A to efficiently undergo autoproteolysis may prevent pathway activation in vivo. To test this hypothesis, we assessed ShhC199A, a form of Shh that is unable to undergo autoproteolytic cleavage (Figure 2B) and is unable to mediate signaling in trans (Roelink et al., 1995). Electroporation of ShhC199A causes an expansion of ventral markers as well as a repression of Pax7 (Figure 5D, H, L). This unanticipated result rejects the hypothesis that the inability of ShhH183A to undergo autoproteolysis is the reason why it is inactive in vivo, and this issue remains unresolved. However, the ability of ShhN-E90A, ShhN-183A, and ShhE90A to induce the Hh response in the developing neural tube provides in vivo support of our finding that Smo can be activated by ShhN without the need to bind to its canonical (co-)receptors.
We stained electroporated neural tubes with the anti-Shh monoclonal antibody 5E1 (Ericson et al., 1996). 5E1 staining pattern on the (left) side of neural tubes overexpressing ShhN or Shh coincide with GFP expressing cells (Figure 3K and 4M). The 5E1 staining of neural tubes expressing ShhN-E90A and in particular ShhE90A is more profuse than neural tubes expressing ShhN or Shh (Figure 4L and 5N). There are two plausible explanations for this observation. 1) The inability of Shh to bind to the co-receptors could greater access to the 5E1 epitope, or 2) inactivation of the putative cannibalistic protease activity of Shh (Rebollido-Rios et al., 2014) might result in higher steady state levels of ShhE90A. The 5E1 staining pattern on neural tubes expressing ShhN-H183A, ShhH183A and Shh-C199A reveals only endogenous Shh expression in the floor plate (Figure 4M, 5O, and 5P). The discrepancy between the effects on patterning and the absence of 5E1 staining of these mutants is possibly caused by poor recognition of the uncleaved ligand, in combination with the disruption of the epitope in the H183Shh mutant. Importantly, the lack of ectopic 5E1 staining in embryos expressing ShhN-H183A and ShhC199A that show ectopic activation of the Hh response, demonstrates that the effect on neural tube patterning is mediated by these Shh mutants as such, and not due to secondary induction of endogenous Shh expression.
The cell-autonomous activity of ShhN is unaffected by alternate C-terminal extensions, but is impeded by the cholesterol modification
The observation that the expression of ShhN, which lacks the C-terminal cholesterol modification, is a much more potent activator of the Hh response than Shh led us to assess the extent to which forms of Shh with varying lipophilic and C-terminal modifications can signal cell-autonomously and in trans. We evaluated fully modified Shh, as well as forms of Shh lacking either the cholesterol moiety (ShhN), the palmitoyl moiety (ShhC25S), or both (ShhN-C25S) (Bumcrot et al., 1995b; Gao et al., 2011; Pepinsky et al., 1998; Porter et al., 1996) (Figure 5A).
The ability of Shh to signal to Ptch1LacZ/LacZ;Ptch2-/- reporter cells in trans (Figure 5A, top diagram) was unaffected by the absence or presence of its lipophilic modifications, with the exception of ShhN-C25S (Figure 5B). The absence of palmitoylation on ShhN results in an inability to signal in trans, as observed before (Chen et al., 2004). ShhN with distinct C-terminal extensions have been shown to be active in embryos before: ShhN::CD4 induces digit duplications when expressed in limb buds (Yang et al., 1997), and ShhN::GFP can partially rescue the genetic loss of Shh (Chamberlain et al., 2008). Both ShhN::CD4 and ShhN::GFP are membrane bound; ShhN::CD4 by its transmembrane domain in CD4, and ShhN::GFP due a cholesterol moiety at the carboxyterminal end of GFP. Both ShhN::CD4 and ShhN::GFP retain the ability to activate the Hh response in trans (Figure 5B). Consistent with previous observations (Roelink et al., 1995), we find that ShhC199A, which is unable to undergo autoproteolytic cleavage, is unable to significantly induce Hh response in trans.
The finding that cells expressing wild type Shh are poorly responsive to their own signal has been reported before (García-Zaragoza et al., 2012), and consistent with the finding that Shh primarily signals in trans (Bailey et al., 2009; Tian et al., 2009). This is in stark contrast to the observation that ShhN is a potent cell-autonomous activator. To evaluate the extent to which the lipophilic modifications of Shh affect its ability to activate the Hh response cell-autonomously, we transfected Ptch1LacZ/LacZ;Ptch2-/- reporter cells with constructs coding for various forms of lipid-modified Shh (Figure 5A, bottom diagram). We find that the palmitoylation of Shh has little effect on its poor ability to induce a cell-autonomous response, unlike its activity in trans (Figure 5C). Similarly, preventing palmitoylation of ShhN (ShhN-C25S) does not affect its ability to induce the Hh response cell-autonomously. Interaction between the palmitate of Shh and Ptch1 has shown to be critical for repressing Ptch1-mediated inhibition (Tukachinsky et al., 2016), and thus the ability of ShhN-C25S to activate the pathway supports the idea that cell-autonomous pathway activation occurs independently of Ptch1/2-mediated inhibition of Smo. This result is further reinforced by the observation that expression of ShhN-C25S causes an upregulation of the Hh response in the developing neural tube (Figure 6, Figure Supplement 1).
The presence of cholesterol on Shh significantly decreases its ability to induce the Hh response cell-autonomously (Figure 6C, 3F, 3I). This decrease in activity is not due to its membrane association, as Shh::CD4 and Shh::GFP are potent cell-autonomous activators of the Hh pathway (Figure 6C). Consistent with the observation that forms of Shh with extraneous C termini are able to induce the Hh response cell-autonomously, we found that the uncleaved ShhC199A mutant also retains the ability to cell-autonomously activate the pathway in Ptch1LacZ/LacZ;Ptch2-/- cells (Figure 6C). These results concur with our observation that ShhC199A can induce the Hh response in vivo (Figure 5D, H, L). It appears that an important function of the cholesterol moiety on Shh is to prevent Shh-expressing cells from strongly upregulating the Hh response cell-autonomously. To test whether the cholesterol modification on Shh alters its cellular distribution, we used the 5E1 antibody to stain transfected Ptch1LacZ/LacZ;Ptch2-/- cells for either wildtype Shh or ShhN. We find that Shh is stereotypically localized to the cell membrane, including filopodial extensions (Figure 6D). ShhN is consistently absent from the cell membrane and filopodial structures, and instead occupies a perinuclear location (Figure 6E). This suggests that the processing and subsequent cholesterol modification of Shh regulates its localization within expressing cells, possibly preventing co-localization with Smo or interaction with other proteins. The cholesterol modification biases against the activation of the Hh response in cells that produce Shh, thereby restricting its effects to non-cell autonomous activation of the Hh pathway. Together these results support our observation that many forms of Shh can activate Smo via a mechanism that does not involve any of the known co-receptors.
Discussion
The canonical model of Smo activation mediated by the Shh-induced release of Ptch1 inhibition is challenged by our observations. Despite the ample evidence that Ptch1 is an efficient inhibitor of Smo (Goodrich et al., 1997), we find that the loss of Ptch1/2 does not inevitably result in maximal Smo activation, and that a variety of Shh mutants can activate the migrational and transcriptional Hh responses in the absence of Ptch1/2 function.
Whether cells lacking Ptch1 or Ptch2 retain sensitivity to Shh in vivo is unclear, however some provocative evidence indicates that they might. The spectrum of tumors observed in Ptch1 null and in Ptch1/2 heterozygous mice (Lee et al., 2006), from basal cell carcinomas to medulloblastomas (Goodrich et al., 1997; Lee et al., 2006; Mao et al., 2006), are commonly found at sites where Shh is expressed; and Shh-induced tumors often arise at the same locations (Beachy et al., 2004). According to the canonical model, the loss of Ptch1/2 function should result in the cell-autonomous activation of the Hh response. The positional overlap between tumors induced by the loss of Ptch function and Shh-induced tumors is compatible with our model that cells lacking Ptch1/2 function nevertheless retain some ability to respond to Shh. Our model is further supported by the observation that Smo is required for the establishment of left/right symmetry in mouse embryos, whereas Ptch1 is not (Zhang et al., 2001b). Pitx2 expression can be induced by Shh (Ryan et al., 1998), and requires Smo as well as Shh and Ihh (Zhang et al., 2001b). Interestingly, Ptch1-/- mutant embryos normally establish asymmetric expression of Pitx2, demonstrating that this Hh/Smo-mediated symmetry-breaking event can occur independently of Ptch1. Analysis of Ptch1-/-;Ptch2-/-;Shh-/-;Ihh-/- compound null embryos would help to resolve the issue as to what extent the loss of Ptch1/2 activity is epistatic to the loss of Shh and Ihh.
In Drosophila embryos, the genetic loss of Ptch is epistatic to the loss of Hh (Bejsovec and Wieschaus, 1993), suggesting that Ptch-independent regulation of Smo does not play a major role during Drosophila development. The incongruity between our data and Drosophila genetics may be explained by auxiliary molecules involved in Drosophila Smo activation that are present in cells lacking Ptch, creating the impression that loss of Ptch-mediated inhibition unequivocally results in Smo activation. However, our model is consistent with the finding that cells in the posterior compartment of the wing imaginal disc retain responsiveness to the Hh ligand independent of Ptch function (Ramirez-Weber et al., 2000), highlighting a mechanism of Smo activation in Drosophila akin to the one we observe in Ptch1LacZ/LacZ;Ptch2-/- cells. It is not terribly surprising that Smo, expressed from a single gene in most vertebrates and invertebrates, would have a complex mechanism of regulation and activation, given that it is responsible for relaying most known Hh responses in the organism.
Our results demonstrate that a wide variety of Shh mutants, many previously characterized as “dead” signals, retain their ability to activate Smo in cell lines lacking the canonical receptors as well as in wildtype cells within developing embryos. Of all the forms of Shh tested, wildtype Shh is unique in that it is unable to induce the cell-autonomous Hh response, while it can still efficiently signal non-cell autonomously. In stark contrast, ShhC199A, which remains unprocessed as a full length precursor, cannot signal to neighboring cells, but is an efficient inducer of the Hh response cell-autonomously. These results indicate that an important function of the unusual processing and modification of Shh is to prevent activation of the Hh response in cells that express the ligand, thus reinforcing strict non-cell autonomous signaling. The ability of ShhC199A to induce the Hh response autonomously came as a surprise, and this switch between cell-autonomous to non-cell autonomous activity upon cleavage has implications for the understanding of diseases associated with mutations in Shh.
Several Shh mutations that cause Holoprosencephaly result in forms of Shh that fail to autoproteolytically cleave, which strongly correlates with their inability to induce the Hh response in trans (Singh et al., 2009; Traiffort et al., 2004). Moreover, many of the otherwise uncharacterized SHH mutants found in patients with Holoprosencephaly are located in the C-terminal domain, and thus possibly affect autoproteolytic cleavage (Roessler et al., 2009). We found that ShhC199A is perfectly capable of inducing the Hh response when expressed in the developing neural tube. It appears that many mutations in the C-terminal domain of SHH could give rise to forms that gain the ability to induce the Hh response cell-autonomously. Heterozygous mutations in the C-terminal domain of SHH can cause holoprosencephaly (Hehr et al., 2010); however, mice heterozygous for a null Shh allele are normal. Two identified Shh mutations that cause Holoprosencephaly change the Cysteine residue at position 198 that is required for Shh processing (homologous to the C199 residue in mice), thus resulting in an obligatory full-length form of Shh (Roessler et al., 2009). Thus, either unprocessed Shh functions as a dominant negative form of Shh, or cell-autonomous activation of the Hh response in cells heterozygous for this mutation contributes to holoprosencephaly. Given our observation that ShhC199A is a potent inducer of the Hh response in vivo, we favor the latter explanation.
The data we present here, along with recently published results from our lab (Roberts et al., 2016) and data from the literature, supports a model in which Ptch1/2 mediates the allosteric inhibition of Smo by means of its antiporter function in the absence of the Shh ligand. Upon the availability of Shh in the extracellular space, Shh can bind to a receptor complex via the ligand-binding domain (L2) of Ptch1/2. We hypothesize that this event represses the allosteric inhibition mediated by Ptch1/2 (Figure 7). The Shh ligand is internalized by the receptor complex, resulting in its co-localization with Smo in an endocytic compartment where it mediates an orthosteric activation of (disinhibited) Smo. It is important to stress that Shh-mediated activation may not occur through direct binding to Smo. Although Smo was initially considered to be the putative receptor for Hh when it was identified in Drosophila, there has been no evidence to support direct binding between Shh and Smo. Many questions about cell-autonomous Hh activation by Shh are left to be explored, however a central question remains: What is the basis of the interaction between Shh and Smo? A multidisciplinary approach of biochemistry, genetics, and bioinformatics may be required to solve this enduring puzzle.
Materials and Methods
Materials
Vismodegib was a gift from Dr. Fred de Sauvage (Genentech). SAG was from EMD Biochemicals. Recombinant ShhN protein was from R&D Systems. Cell Tracker Green CMFDA was from Invitrogen.
Electroporations
Hamburger-Hamilton (HH) stage 10 Gallus gallus embryos were electroporated caudally in the developing neural tube using standard procedures (Meyer and Roelink, 2003). Embryos were incubated for another 48 h following electroporation, dissected, fixed in 4% PFA, mounted in Tissue-Tek OCT Compound (Sakura) and sectioned.
Embryoid Body differentiation
mESCs were neuralized and differentiated into embryoid bodies (NEBs) using established procedures (Wichterle et al., 2002). NEBs were harvested after 5 days in culture, fixed, and stained for Nkx2.2 and Olig2 (Kawakami et al., 1997). NEBs were mounted in Fluoromount-G (Southern Biotech) and imaged.
Immunofluorescence
Antibodies for mouse Pax7 (1:10), Mnr2 (1:100), Nkx2.2 (1:10), Shh (5E1, 1:20) were from the Developmental Studies Hybridoma Bank. The Rabbit α-GFP (1:1000) antibody was from Invitrogen, and the Goat α-hOlig2 (1:100) antibody was from R&D Systems. The mouse α-acetylated tubulin (1:200) was from Sigma Aldrich. Alexa488 and Alexa568 secondary antibodies (1:1000) were from Invitrogen. Nuclei were stained with DAPI (Invitrogen).
DNA Constructs
The Gli-Luciferase reporter and the Renilla control were gifts from Dr. H. Sasaki (Sasaki et al., 1997). Ptch1 was a gift from Dr. Scott (Stanford University, CA, USA). Ptch1-Δloop2 was a gift from Dr. Thomas Jessell (Columbia University, NY, USA). Ptch1 channel mutants were previously described (Alfaro et al., 2014). SmoΔCRD was a gift from J. Reiter (Aanstad et. al., 2009). SmoM2 was from Genentech (F. de Sauvage). The following mutations were created using Quikchange mutagenesis (Stratagene): Shh-E90A, Shh-H183A, ShhN-E90A, ShhN-H183A, SmoΔCRD-CLD, SmoL112DW113Y. Shh::GFP and Shh::CD4 were gifts from Dr. Andrew McMahon (University of Southern California, CA, USA). Dhh and Ihh were gifts from Charles Emerson Jr. (University of Massachusetts Medical School, MA, USA). DhhN and IhhN were made by site directed mutagenesis of C199 (Dhh) and C203 (Ihh) to a stop codon. ShhC199A was previously described (Roelink et al., 1995). Gli3PHS was previously described (Meyer et. al., 2003).
Genome Editing
TALEN constructs targeting the first exon of mouse Cdo and Gas1 were designed and cloned into the pCTIG expression vectors containing IRES puromycin and IRES hygromycin selectable markers (Cermak et al., 2011). The following repeat variable domain sequences were generated: Cdo, 5’ TALEN: NN HD NI NG HD HD NI NN NI HD HD NG HD NN NN; 3’ TALEN: HD NI HD NI NI NN NI NI HD NI NG NI HD NI NN; Gas1, 5’ TALEN: NN NI NN NN NI HD NN HD HD HD NI NG NN HD HD; 3’ TALEN: NN NN NI NI NI NI NN NG NG NG NN NG HD HD NN NI. Two CRISPR constructs targeting a double strand break flanking the first exon of mouse Boc were cloned into pSpCas9 vector with an IRES puromycin selectable marker (Ran et al., 2013). The Boc CRISPRs targeted the following forward genomic sequences (PAM sequences underlined): Upstream of first exon 5’ CCTGTCCTCGCTGTTGGTCCCTA 3’; Downstream of first exon 5’ CCCACAGACTCGCTGAAGAGCTC 3’. Ptch1LacZ/LacZ;Ptch2-/-;Shh-/- mouse embryonic stem cells (Roberts et al., 2016) were plated at a density of 1.0x106 on 6-well plates and transfected with 6 genome editing plasmids the following day. One day after transfection, selective ES medium (100 μg/mL hygromycin and 0.5 μg/mL puromycin) was added for 4 days. Selective medium was removed and surviving mESC colonies were isolated, expanded and genotyped by sequence PCR products spanning TALEN and CRISPR-binding sites.
Genotyping
PCR screening was performed on cell lysates using primers flanking the TALEN or CRISPR binding sites for the Boc, Cdo, and Gas1 loci. Boc, (5’) CATCTAACAGCGTTGTCCAACAATG and (3’) CAAGGTGGTATTGTCCGGATC; Cdo, (5’) CACTTCAGTGTGATCTCCAG and (3’) CCTTGAACTCACAGAGATTCG; Gas1, (5’) ATGCCAGAGCTGCGAAGTGCTA and (3’) AGCGCCTGCCAGCAGATGAG. PCR products were sequenced. Samples with signals indicative of INDEL mutations were cloned into the pCR-BluntII vector using the Zero Blunt TOPO PCR cloning kit (Invitrogen) and sequenced to confirm allele sequences. A Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- mESC clone was identified harboring a 50 bp deletion in Cdo exon 1, a heteroallelic 480 bp insertion and a 200 bp deletion in Gas1 exon1 resulting in a premature stop codon in the reading frame, and a 450 bp deletion of Boc exon 1.
Cell Culture
Ptch1LacZ/LacZ;Ptch2-/-, Ptch1LacZ/LacZ;Ptch2-/-;Smo-/-, Ptch1-/-;Disp1-/-;Shh-/-, and Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- fibroblasts were obtained by plating mESCs at a density of 8.0×105 cells in 6-well plates and transfected with the large T antigen from the SV40 virus (Gökhan et al., 1998) in ES medium. Cells were then switched to DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS) without LIF. Ptch1+/LacZ and Ptch1LacZ/LacZ fibroblasts (gifts from Dr. Scott) were cultured in DMEM supplemented with 10% FBS (Invitrogen) and maintained under standard conditions. Identity of these lines was confirmed by the presence of the LacZ recombination in the Ptch1 locus, the presence of 40 chromosomes per cell, and mouse-specific DNA sequences of the edited genes. mESC lines were maintained using standard conditions in dishes coated with gelatin, without feeder cells. Cells were routinely tested for Mycoplasm by Hoechst stain, and grown in the presence of tetracycline and gentamycin at regular intervals. Cultures with visible Mycoplasma infection were discarded. None of the cell lines used in this study is listed in the Database of Cross-Contaminated or Misidentified Cell Lines. Fibroblast-like lines derived from the mESCs were re-sequenced at the edited loci to confirm their identity.
Transfection
Cells were transiently transfected for 24h at 80-90% confluency using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s protocol.
Modified Boyden Chamber Migration Assay
Cell migration assays were performed as previously described (Bijlsma et. al. 2007). Cells were labeled with 10 μM CellTracker Green (Invitrogen) in DMEM for one hour. The well compartments were set up with the specified chemoattractant (ShhN .75 μg/ML (resuspended in 0.1%BSA in PBS), 10% FCS, or no attractant (plus 0.1%BSA in PBS) and pre-warmed at 37°C. Cells were then detached with 5mM EDTA and resuspended in DMEM without phenol red and supplemented with 50mM HEPES. Cells were transferred into FluoroBlok Transwell inserts (BD Falcon) at 5.0×104 cells per insert. GFP-spectrum fluorescence in the bottom compartment was measured every 2 min for 99 cycles (approximately 3 hours), after which background fluorescence (medium without cells) and a no-attractant control was subtracted from each time point. Starting points of migration were set to 0.
Gli-Luciferase Assay
Fibroblasts were plated at a density of 3×104 in 24 well plates and transfected with Gli-Luciferase, CMV-Renilla (control plasmid), and specified plasmids 24 hours after plating. Cells were grown to confluency and then switched to low serum medium (0.5% FBS) alone or with specified concentrations of Vismodegib or SAG. After 24 hours, cells were lysed and the luciferase activity in lysates was measured using the Dual Luciferase Reporter Assay System (Promega). Raw Luciferase values were normalized against Renilla values for each biological replicate to control against variation in transfection efficiency. Individual luciferase/renilla values were then normalized against the mock control average for each experiment.
LacZ Assay
Fibroblasts were plated at a density of 3×104 in 24 well plates and transfected with plasmids 24 hours after plating (or remained untransfected for SAG experiments). Fibroblasts were grown to confluency and then switched to a low serum medium (0.5% FBS) alone or with specified concentrations of Vismodegib or SAG. After 24 hours, cells were lysed and lysates were analyzed using the Galacto-Light™ chemiluminescence kit (Applied Biosciences) for level of LacZ expression. Raw chemiluminescence values were normalized against total protein for each biological replicate. Protein concentration was determined with a Bradford assay using the Bio-Rad Protein Assay Dye Reagent.
Co-Culture Assay
Ptch1LacZ/LacZ;Ptch2-/- reporter cells were plated at a density of 1.4×105 in 6-well plates and transfected with Gli-Luciferase, CMV-Renilla, and with or without a variant of Ptch1 24 hours after plating. Ptch1LacZ/LacZ;Ptch2-/- signaling cells were plated similarly and transfected with a variant of Shh 24 hours after plating. Cells were then trypsinized and plated in 24-well plates in specified combinations (1.5×104 of each type) 24 hours after transfections. Cells were grown to confluency and switched to low serum medium (0.5% FBS) for 16 hours before assaying for luciferase activity.
Western Blots
Ptch1LacZ/LacZ;Ptch2-/- cells were transfected with Shh mutants as indicated. 48 hours after transfection, Ptch1LacZ/LacZ;Ptch2-/- cells were rinsed with PBS and lysed with RIPA buffer (150 mM NaCl, 50 mM Tris-Hcl, 1% Igepal, 0.5% Sodium Deoxycholate, and protease inhibitors) for 30 min on ice. Protein lysate was cleared by centrifugation at 13,000g for 30 min at 4 °C. 20 μg of each sample was run on a 15% SDS-PAGE gel and transferred to a 0.45 micron nitrocellulose membrane. Membranes were blocked with 5% milk in Tris-buffered saline with 0.1% Tween-20 (TBS-T) and incubated with a rabbit polyclonal anti-Shh antibody (H160; Santa Cruz Biotechnology) at 1:250. A goat anti-rabbit HRP-conjugated secondary antibody (Biorad) was used at 1:10000.
Author contributions
CC performed all experiments. HR and CC designed the experiments and wrote the manuscript.
Acknowledgements
This work was supported by NIH grants R01GM097035 and 1R01GM117090 to HR. Vismodegib was a gift from Dr. de Sauvage (Genentech), Dhh and Ihh were a gift from Dr. Charles P. Emerson III (U. Mass. Med. School). Dr. A. Alfaro cloned the Shh binding mutants (E90A, H183A) and made the initial observation that they activate the Hh pathway in vivo. We thank Dr. M. Barro for her help with Western blotting, Dr. M. F. Bijlsma (AMC Amsterdam) for his help with the migration assays, and Dr. B. Roberts (Allen Institute for Cell Science) for his comments and advice. We also thank Fatma Ozguc for help with genome editing of the Ptch1LacZ/LacZ;Ptch2-/-;Boc-/-;Cdo-/-;Gas1-/-;Shh-/- mESC line.
Footnotes
roelink{at}berkeley.edu