Jin Woo Yoo

Harvard College ‘12

Yu Wang, Andrew P. McMahon

Department of Stem Cell and Regenerative Biology, Department of Molecular and Cellular Biology,
Department of Chemistry and Chemical Biology, Harvard Stem Cell Institute

Abstract

The activation of Hedgehog (Hh) signaling has been shown to induce translocation of core Hh components to the primary cilium (PC). Notably, Smoothened (Smo), a key mediator in the pathway, rapidly accumulates at the PC in response to Hh stimulation. While the ciliary accumulation of Smo has been well-noted, the current knowledge of PC biology in the context of Hh signaling is primitive at best, limited to an on-and-off type of binary understanding. Using an in vitro approach coupled with quantitative fluorescence imaging, we identified at least four discrete patterns of Smo distribution that fell into well-separated regions within the PC, in which a tip-based ciliary localization of Smo correlated with the activation of Hh transcription in the nucleus. In parallel with the Smo localization study, we also found evidence for ongoing, ligand-independent Hh signaling that potentially constitutes an internal, auto-regulatory loop maintaining Hh responsiveness. Collectively, our findings suggest that in an unstimulated cell, Smo continually traffics in and out of the PC, and upon stimulation, Smo accumulates at the PC in a tightly regulated, sequential fashion, in which its accumulation at the ciliary tip is necessary for the relay of Hh signaling.

Introduction

The Hedgehog (Hh) signaling pathway, a key regulator of embryonic development, has critical roles in organizing pattern formation in the embryonic limb, neural tube, and other structures (Ingham and McMahon, 2001). Pathway dysfunction during development can cause severe birth defects and is often lethal. For example, insufficient Hh signaling can cause holoprosencephaly, or the failure of the brain to properly segregate into two halves. Conversely, excessive Hh signaling commonly results in polydactyly, the formation of supernumerary digits (McMahon et al., 2003). Even after development, the Hh pathway continues to have important biological roles in tissue maintenance and repair. Numerous studies have established that deregulation of Hh pathway is associated with the development of various cancers, including basal cell carcinoma, a type of skin cancer, and medulloblastoma, a childhood brain tumor (Xie, 1998; Romer, 2004).

The fundamental importance of Hh signaling in development, physiology, and disease has motivated extensive study of the pathway. Studies of Hh over the past 30 years have elucidated many molecular components of the pathway as well as their basic functions (Barakat, 2010, Hooper, 2005). Briefly, Hh ligands activate the pathway by binding to its receptor, Patched1 (Ptch1), thereby relieving Ptch1-mediated inhibition of Smoothened (Smo), a 7-pass transmembrane protein essential for active Hh signaling. Upon binding of ligand to Ptch1, Smo subsequently becomes activated and ultimately leads to an inhibition of repressor and generation of activator forms of Ci/Gli proteins. In Drosophila, all Hh-dependent transcriptional regulation is carried out by Ci, its function is subdivided among three Gli proteins in vertebrates (Gli1-3). Gli1 is nonessential and functions exclusively as a transcriptional activator dependent upon the activity of Gli2 and Gli3 (Park, 2000). Gli2 and Gli3 can exist in either full-length or proteolytically processed repressor forms, although Gli2 primarily acts as an activator and Gli3 as a repressor (Bai, 2002; Persson, 2002; Wang, 2000). The ratio of the activator and repressor functions of Gli proteins is precisely determined by the status of Hh pathway. The activation of Gli factors ultimately results in the transcriptional up-regulation of Hh target genes, which include Ptch1 and Gli1, that act as feedback components of the pathway (Jiang, 2008).

Acting in between Smo and Gli proteins is Suppressor of Fused (SuFu), a key negative regulator of the Hh pathway. Loss of function mutations affecting the gene encoding SuFu result in maximally activated Hh pathway independent of ligand (Ding, 1999; Kogerman, 1999; Methot and Basler, 2000). In the absence of Hh ligand, SuFu physically forms a complex with Gli proteins in the cytoplasm and prevents their nuclear entry (Kogerman, 1999; Pearse, 1999). Upon Hh stimulation, the defined Gli-SuFu complex becomes dissociated, permitting Gli nuclear entry and the transcription of Hh target genes (Tukachinsky, 2010).

A key feature of Hh signaling in vertebrates versus Drospohila is the essential role of primary cilium (PC), a tubulin-based cell extension present on most non-dividing cells (Goetz, 2010). Initial evidence for a major role of PC in Hh came from genetic screens that identified a phenotypic similarity in mutations affecting intraflagellar transport proteins (IFTs) and those in genes encoding Hh components. A phenotypic similarity with Hh mutants is evident in mutations affecting motor proteins involved in ciliary transport or ciliary structure (Eggenschwiler, 2007; Huangfu, 2003). Given the dual role for the centriole in generating the mitotic spindle in replicating cells and the primary cilium in quiescent cells, these findings explain the requirement for high cell density or serum starvation to mount a Hh response in vitro (Pugacheva, 2007).

Several major components of the Hh pathway localize to the PC, including Ptch1, Smo, SuFu, as well as Gli transcription factors (Haycraft, 2005; Kiprilov, 2008; Rohatgi, 2007; Tran, 2008). In an unstimulated cell, Ptch1 constitutively localizes near the basal body of the PC. Upon engagement with Shh, Ptch1 localization is lost, and Smo translocates to the PC (Figure 1a, 1b) (Rohatgi, 2007). Translocation of Smo to the primary cilium is essential for Hh pathway activation (Corbit, 2005; Wilson, 2009). Mutations in Smo that confer constitutive pathway activity (SmoA1) also result in constitutive accumulation of Smo at the PC; conversely, mutations that abrogate ciliary accumulation of Smo (CLDSmo) render the protein incapable of activating the pathway in the presence of intact PC (Corbit, 2005). Interestingly, Smo translocation to the PC can be forced by both small molecule agonists as well as antagonists. Notably, cyclopamine (Cyc) induces Smo accumulation at the PC yet acts as an overall Hh antagonist (Wang, 2009). Thus, Smo may assume multiple active or inactive conformations that influence its status at the PC and its capacity to relay Hh signaling to the downstream components (Wilson, 2009).

Downstream of Smo, SuFu and Gli proteins co-localize at the distal tip of the PC, consistent with the biochemical data showing their direct physical interaction (Huangfu, 2005). In the absence of Shh ligand, Gli and SuFu proteins appear to traffic to the PC as a complex at a low level independently of Smo, resulting in a weak accumulation at the ciliary tip. Hh stimulation through active Smo results in the recruitment of Gli-SuFu complexes to the PC and causes their ciliary levels to rapidly increase. This then leads to the dissociation of Gli-SuFu complexes, and once Gli proteins are relieved from SuFu suppression, Gli proteins are free to enter the nucleus and regulate their transcriptional targets (Figure 1b, 1c) (Tukachinsky, 2010; Zeng, 2010).

Given that Smo translocation to the PC is required for pathway activation, Smo has emerged as the predominant therapeutic target of anti-Hh drug discovery efforts. Given that the ciliary accumulation of Smo upon Hh activation is readily noticeable, novel imaging-based drug screens have been developed, in which compounds are screened based on their ability to prevent Smo translocation to the PC. However, while the ciliary trafficking of Smo has been well documented as a general phenomenon, our understanding of this process is rudimentary at best; ciliary accumulation has been reduced to the function of a simple light switch that merely informs us of whether something is turned on or off. How does Smo become activated once it is relieved from Ptch1 inhibition? What changes are made since its arrival at the PC? What is the functional significance of the difference observed in the ciliary distribution of Smo (uniform) versus Gli or Sufu (distal tip-based)?

Here, we have attempted to examine these questions via a temporal and spatial analysis of Smo at the PC.

Figure 1. Schematic construction of Hh signaling at the PC. (a) In the absence of Shh ligand, Ptch1 localizes to the base of PC, preventing Smo ciliary ac- cumulation. Gli and SuFu traffic to the primary cilium at a low level independently of Smo. (b) Upon engagement with Shh, Ptch1 localization at the base of the PC is lost. Smo is then relieved from Ptch1-mediated repression, and the active form of Smo accumulates at the PC, causing an increase in the ciliary levels of Gli-SuFu complex. (c) Upon activation of Smo, Gli-SuFu complex subsequently becomes dissociated, freeing Gli proteins to enter the nucleus and regulate their transcriptional targets, including Ptch1 and Gli1 that serve as feedback components.
Figure 1. Schematic construction of Hh signaling at the PC. (a) In the absence of Shh ligand, Ptch1 localizes to the base of PC, preventing Smo ciliary ac- cumulation. Gli and SuFu traffic to the primary cilium at a low level independently of Smo. (b) Upon engagement with Shh, Ptch1 localization at the base of the PC is lost. Smo is then relieved from Ptch1-mediated repression, and the active form of Smo accumulates at the PC, causing an increase in the ciliary levels of Gli-SuFu complex. (c) Upon activation of Smo, Gli-SuFu complex subsequently becomes dissociated, freeing Gli proteins to enter the nucleus and regulate their transcriptional targets, including Ptch1 and Gli1 that serve as feedback components.

Results

Smo::EGFP is behaviorally and functionally similar to endogenous Smo

For the visualization of Smo, an EGFP tagged form of human Smo was introduced to Hh-responsive NIH3T3 cells. An Inversin (Ivs)::tagRFP-T expression cassette was also introduced as a Hh independent, constitutive PC marker that is necessary for the identification of intact PC as well as quantitative image analysis of PC (Watanabe, 2003). A caveat to fluorescently tagging proteins is the possibility that tagging may alter the endogenous behavior or the function of the protein of interest. To rule out this possibility, a comparative analysis of Smo::EGFP versus endogenous Smo detected by antibody staining was performed (Figure 2a-b). Smo::EGFP / Ivs::tagRFP-T transgenic cells and their parental, wild-type NIH3T3 cells were seeded side by side on a cover-slip bottom chamber slide. Cells were cultured to confluency and serum-starved for 24 hours to promote PC assembly. A comprehensive set of treatments (from now on referred to as standard treatments), selected based on their well-known effects on Hh signaling, included: Shh, Smo agonist (SAG), Cyc (Smo inhibitor associated with ciliary accumulation of Smo), Cyc+Shh, SANT-1 (a potent Smo antagonist blocking ciliary accumulation), and SANT-1+Shh. After treatment for 24 hours, NIH3T3 cells were antibody stained for Smo and acetylated tubulin for independent labeling of PC.

The percentage of Smo positive PC was determined by identifying intact PC through Ivs::tagRFP-T or antibody staining for acetylated tubulin, and determining the status of Smo at each of the defined PC. No significant differences in the percentage of Smo positive PC were observed for Smo::EGFP versus endogenous Smo, suggesting that Smo::EGFP mirrors the trafficking behavior of endogenous Smo. As expected, Shh, SAG, and Cyc were able to induce robust ciliary translocation of both Smo::EGFP and endogenous Smo. Furthermore, SANT-1 was able to suppress the ligand-mediated ciliary translocation of both Smo::EGFP and endogenous Smo. Finally, to determine whether Smo::EGFP is functionally equivalent to endogenous Smo, Gli-Luciferase reporter assay was performed in order to measure Hh transcriptional response under standard treatments (Figure 2c). No significant differences in the levels of Hh transcriptional response were observed for Smo::EGFP / Ivs::tagRFP-T transgenic cells versus NIH3T3 cells, suggesting their functional equivalence. A more definitive functional assay of the Smo::EGFP / Ivs::tagRFP-T cell line was performed in a previous study, in which the expression of Smo::EGFP was shown to rescue Hh response in cells after shRNA-mediated knockdown of endogenous Smo (Wang, 2009).

Figure 2. Smo::EGFP is behaviorally and functionally similar to endog- enous Smo. (a) Comparative imaging analysis of Smo::EGFP and endogenous Smo confirms that Smo::EGFP mirrors the movement of endogenous Smo under standard treatments. Standard treatments included: Shh (20nM), SAG (1μM), Cyc (5μM), Cyc+Shh, SANT-1 (1μM), and SANT-1+Shh. The concen- tration for each treatment was conserved throughout the study. NIH3T3 cells were immunostained for Smo and acetylated tubulin (AcTub) to label the PC. Scale bar: 5μm. (b) Quantification of the images produced in (a). (c) Gli-Lu- ciferase reporter assay comparing the transcriptional response in Smo::EGFP / Ivs::tagRFP-T cells versus NIH3T3 cells. No significant differences in Hh re- sponse between the two cell lines could be observed across the standard treat- ments. Mean (±S.D.) was calculated from six replicate samples.
Figure 2. Smo::EGFP is behaviorally and functionally similar to endog- enous Smo. (a) Comparative imaging analysis of Smo::EGFP and endogenous Smo confirms that Smo::EGFP mirrors the movement of endogenous Smo under standard treatments. Standard treatments included: Shh (20nM), SAG (1μM), Cyc (5μM), Cyc+Shh, SANT-1 (1μM), and SANT-1+Shh. The concen- tration for each treatment was conserved throughout the study. NIH3T3 cells were immunostained for Smo and acetylated tubulin (AcTub) to label the PC. Scale bar: 5μm. (b) Quantification of the images produced in (a). (c) Gli-Lu- ciferase reporter assay comparing the transcriptional response in Smo::EGFP / Ivs::tagRFP-T cells versus NIH3T3 cells. No significant differences in Hh re- sponse between the two cell lines could be observed across the standard treat- ments. Mean (±S.D.) was calculated from six replicate samples.

Kinetics of Smo ciliary trafficking

To gauge the kinetics of Smo trafficking to the PC, Smo::EGFP / Ivs::tagRFP-T cells were seeded in a 384-well imaging plate, cultured to confluency, serum-starved for 24 hours, and then induced with Shh, SAG, or Cyc over a series of time points up to 24 hours. The plate was then imaged, and all subsequent images were quantitatively analyzed based on a custom-designed multi-parametric script that was developed by Y. Wang (Harvard University, Cambridge, MA) and L. Davidow (Harvard University, Cambridge, MA). To quantify the levels of Smo at the PC, the mean EGFP intensity in the PC and fraction of EGFP positive PC were measured, both parameters normalized to the absolute cilia count as determined by Ivs positive structures. Results produced by the script-based analysis were consistent with the expected Smo trafficking behaviors under standard treatments, demonstrating the script’s effectiveness in studying the kinetics of Smo ciliary trafficking (Figure 3a).

Smo accumulation in the PC could be detected as early as 30 minutes after the addition of Shh; the mean Smo intensity and the fraction of Smo positive PC rapidly increased over the time course of analysis such that at 3 hours, approximately 75 percent of the cells displayed a significant level of Smo at the primary cilium (Figure 3b). No further significant increase in either mean ciliary Smo intensity or the fraction of Smo positive PC was observed, suggesting that maximal Smo ciliary accumulation occured around 3 hours after ligand-mediated Hh stimulation. Treatment with SAG had comparable kinetics; the first significant Smo ciliary accumulation was evident by 20 minutes (Figure 3c). One notable difference in pathway activation was that SAG treated cells had significantly higher mean ciliary Smo intensity than Shh treated cells, probably due to the greater potency of SAG in activating Smo and the inability of ligand-directed feedback mechanisms to attenuate ligand-independent signaling activity. Treatment with Cyc displayed the slowest kinetics, resulting in the first significant Smo ciliary accumulation at 40 minutes (Figure 3d). The delayed kinetics of Smo ciliary translocation in response to Cyc may indicate a qualitative difference in active versus inactive conformations of Smo. However, further work is needed to consider other possibilities, such as targeted degradation that could account for the differences observed.

Figure 3. Kinetics of Smo ciliary trafficking. (a) Assessment of Smo ciliary accumulation under standard treatments. All measurements were quadrupli- cated, with several hundred cells per sample. Image quantifications were pro- duced by a custom-designed script. Results are consistent with the reported findings as well as the manual quantifications produced in Figure 2b, confirm- ing effectiveness of the script-based analysis. (b-d) Assessment of the kinetics of Smo ciliary trafficking when treated with (b) Shh, (c) SAG, and (d) Cyc. All measurements and image quantifications were produced in the same manner as in (a). * indicates statistical significance in a t test comparing to the un- treated control (*p<0.05, **p<0.01).
Figure 3. Kinetics of Smo ciliary trafficking. (a) Assessment of Smo ciliary accumulation under standard treatments. All measurements were quadrupli- cated, with several hundred cells per sample. Image quantifications were pro- duced by a custom-designed script. Results are consistent with the reported findings as well as the manual quantifications produced in Figure 2b, confirm- ing effectiveness of the script-based analysis. (b-d) Assessment of the kinetics of Smo ciliary trafficking when treated with (b) Shh, (c) SAG, and (d) Cyc. All measurements and image quantifications were produced in the same manner as in (a). * indicates statistical significance in a t test comparing to the un- treated control (*p<0.05, **p<0.01).

Evidence for auto-regulatory, background Hh signaling

Interestingly, we noted that 10 percent to 20 percent of cells consistently accumulated Smo at the PC in the absence of any Hh stimulation, which led to the speculation that there might be trace quantities of Shh ligand in the serum that was activating the Hh pathway (Figure 3a). However, no significant changes in the levels of ciliary Smo were observed when the serum concentration was increased from 0.5% to 10% serum relative to the blank DMEM control (Figure 4a). To conclusively rule out the possibility of serum containing trace quantities of Shh ligand, cells were treated with anti-Shh antibody (5E1) in conjunction with the standard treatments to inactivate any potential Shh ligand in the serum (Figure 4b). Consistent with the negative serum titration data, 5E1 could not suppress the basal level of Smo accumulation at the PC, but 5E1 suppressed the effect of exogenous Shh addition to cultured cells. In contrast, cells treated with SANT-1 displayed a significant reduction in both the mean ciliary Smo intensity and the fraction of Smo positive PC relative to untreated or 5E1 treated cells, suggesting that the basal level of Smo ciliary accumulation is not an artifact of background noise produced by the assay but a real event that depends on Smo activity.

Next, we asked if the basal level of Smo ciliary accumulation translated to a difference in Hh transcriptional response. Surprisingly, no significant difference in the relative Gli-Luciferase signal could be observed for 5E1 or SANT-1 treated cells relative to the untreated control (Figure 4c). A potential caveat of Gli-Luciferase reporter assay is the lack of sensitivity that is required to detect small differences in Hh transcriptional response. (Because the Gli-Luciferase reporter system relies on high expression levels of transgenes, the assay is subject to some background noise that may be sufficient to blanket small, biologically significant differences.) Hence, the experiment was repeated using Gli1 mRNA fluorescence in situ hybridization (FISH) that provides a more sensitive transcriptional assay, achieving a level of precision able to detect a single strand of mRNA. Since Gli1 is a weak transcriptional activator as well as a Hh target gene, it serves as a reliable reporter for Hh transcriptional activity. GADPH, a typical housekeeping gene, was used as a positive control. In contrast to the negative Gli-Luciferase data, a significant reduction in the levels of Gli1 mRNA transcripts was observed in cells treated with SANT-1 relative to the untreated control (Figure 5a, 5b). Collectively, these findings provide evidence for ligand-independent, low level Hh signaling that is likely mediated by the continuous cycling of Smo in and out of the PC.

Figure 4. Evidence for auto-regulatory, background Hh signaling. (a) Increasing the concentration of fetal calf serum (FCS) in the media does not induce Smo accumulation at the PC. (b) SANT-1 can suppress the basal level of Smo accumulation at the PC that is observed in unstimulated cells, but not 5E1 (1μg/ml), suggesting that the possibility of serum containing Shh ligand is not likely. Anti-HA antibody (1μg/ml) was used as an IgG control; at the given concentration, Shh induction was not affected, confirming the absence of non-specific effects due to the antibody treatment. (a-b) All measurements were quadruplicated, with several hundred cells per sample. Image quantifica- tions were produced by a custom-designed script. * indicates statistical signifi- cance in a t test (**p<0.01, ****p<0.0001). (c) Gli-Luciferase reporter assay of NIH3T3 cells. Suppression of background Hh signaling by Smo antagonists could not be detected in this assay due to the assay’s intrinsic detection limit. Mean (±S.D.) was calculated from six replicate samples.
Figure 4. Evidence for auto-regulatory, background Hh signaling. (a)
Increasing the concentration of fetal calf serum (FCS) in the media does not induce Smo accumulation at the PC. (b) SANT-1 can suppress the basal level of Smo accumulation at the PC that is observed in unstimulated cells, but not 5E1 (1μg/ml), suggesting that the possibility of serum containing Shh ligand is not likely. Anti-HA antibody (1μg/ml) was used as an IgG control; at the given concentration, Shh induction was not affected, confirming the absence of non-specific effects due to the antibody treatment. (a-b) All measurements were quadruplicated, with several hundred cells per sample. Image quantifica- tions were produced by a custom-designed script. * indicates statistical signifi- cance in a t test (**p<0.01, ****p<0.0001). (c) Gli-Luciferase reporter assay of NIH3T3 cells. Suppression of background Hh signaling by Smo antagonists could not be detected in this assay due to the assay’s intrinsic detection limit. Mean (±S.D.) was calculated from six replicate samples.
Figure 5. Background Hh signaling can be suppressed by SANT-1 but not by 5E1. (a) Gli1 mRNA FISH assay of NIH3T3 cells. Treatment with Cyc or SANT-1 can effectively suppress the background levels of Gli1 mRNA tran- scripts observed in untreated cells, but not 5E1. Each green dot represents an mRNA transcript. GAPDH was used as a positive control for the assay. Treat- ment with 6-Dichlorobenzimidazole riboside (DRB) (100μM; 3h), a general transcription inhibitor, can suppress both Gli1 and GAPDH transcript levels, confirming the assay’s specificity for mRNA transcripts. Scale bar: 5μm.(b) Quantification of images produced in (a). Measurements were quadruplicated, with several hundred cells per sample. mRNA count was produced by measur- ing the total number of green dots normalized by the total cell count. * in- dicates statistical significance in a t test comparing to the untreated control (**p<0.01, ***p<0.001).
Figure 5. Background Hh signaling can be suppressed by SANT-1 but not by 5E1. (a) Gli1 mRNA FISH assay of NIH3T3 cells. Treatment with Cyc or SANT-1 can effectively suppress the background levels of Gli1 mRNA tran- scripts observed in untreated cells, but not 5E1. Each green dot represents an mRNA transcript. GAPDH was used as a positive control for the assay. Treat- ment with 6-Dichlorobenzimidazole riboside (DRB) (100μM; 3h), a general transcription inhibitor, can suppress both Gli1 and GAPDH transcript levels, confirming the assay’s specificity for mRNA transcripts. Scale bar: 5μm.(b) Quantification of images produced in (a). Measurements were quadruplicated, with several hundred cells per sample. mRNA count was produced by measur- ing the total number of green dots normalized by the total cell count. * in- dicates statistical significance in a t test comparing to the untreated control (**p<0.01, ***p<0.001).

Smo fills the PC in a tightly-regulated and sequential fashion

Contrary to the conventional notion that Smo fills the PC uniformly, upon closer observation, we observed an array of strikingly discrete patterns of Smo distribution at the PC. At least four discrete patterns of Smo distribution could be observed along the proximal (basal body) to distal (cilary tip) axis of PC organization and growth: [a] proximal based accumulation, [b] distal based accumulation, [c] dumbbell-shaped accumulation, and finally [d] uniform accumulation throughout the entire length of PC (Figure 6a-6i). Cells were immunostained for γ-tubulin, a basal body component, in order to orient the ciliary trajectory from the cell surface. Proximal based accumulation of Smo consisted of a concentrated aggregate of Smo near the base the PC, followed by a more faint shadow of Smo along the extension of the primary cilium (Figure 6a, 6e), while distal based accumulation showed the opposite pattern (Figure 6b, 6f). Dumbbell-shaped accumulation reflected the sum of the first two patterns, appearing as two intense aggregates of Smo at both ends of the PC (Figure 6c, 6g). Lastly, uniform accumulation was the predominantly observed pattern, which consisted of high levels of Smo filling the entire PC, more or less homogeneously (Figure 6d, 6h). When the plot profiles for each representative pattern were overlaid, the four discrete patterns of Smo distribution fell into well-defined, separated compartments along the PC (Figure 6i).

Smo accumulation was scored in each of these four categories during the course of a Shh-directed signaling response over 3 hours of Hh activation (Figure 6j). The highest frequency of proximal based accumulation occurred at early time points, and gradually decreased after 1 hour of Shh induction. Conversely, the highest frequency of uniform accumulation occurred after 2 hours of Shh induction. Distal based and dumbbell-shaped patterns of accumulation were observed sporadically in between the peaks of a proximal based and uniform accumulation of Smo. At the end of 3 hours of Shh induction, a uniform accumulation throughout the PC was the predominant pattern of Smo ciliary distribution. Collectively, these results suggest two important conclusions. First, the restricted patterns of Smo distribution at the PC suggest a compartmentalized response by Smo, accumulating at the PC in a tightly regulated and sequential fashion. Secondly, the activation of Smo appears to spatially correlate with its accumulation at the ciliary tip. It is likely that accumulation at the ciliary tip facilitates the interaction of Smo with downstream components of the Hh pathway that also localize at the ciliary tip, including SuFu and Gli2 (Figure 7a, 7b).

Figure 6. Smo assumes at least four discrete patterns of distribution in the PC. (a-d) Representative images of the four discrete patterns of Smo distribution in the order of (a) proximal based, (b) distal based, (c) dumbbell- shaped, and (d) uniform accumulation of Smo. Cells were immunostained for γ-tubulin to indicate the basal body. Dotted white lines were drawn to demar- cate the entire length of the PC on the basis of overexposed, high-contrast im- age settings. Scale bar: 2μm.(e-h) Plot profiles of the images produced in (a). Plot profiles were generated by quantifying the EGFP intensity along the mid- axis of each PC using ImageJ. Each sample was quadruplicated. (i) Overlay of the plot profiles produced in (e-h) highlights well-defined, separated regions of Smo accumulation within the PC. (j) Progression from a proximal based accumulation to a uniform accumulation of Smo throughout the PC correlates with Hh pathway activation.
Figure 6. Smo assumes at least four discrete patterns of distribution in the PC. (a-d) Representative images of the four discrete patterns of Smo distribution in the order of (a) proximal based, (b) distal based, (c) dumbbell- shaped, and (d) uniform accumulation of Smo. Cells were immunostained for γ-tubulin to indicate the basal body. Dotted white lines were drawn to demar- cate the entire length of the PC on the basis of overexposed, high-contrast im- age settings. Scale bar: 2μm.(e-h) Plot profiles of the images produced in (a). Plot profiles were generated by quantifying the EGFP intensity along the mid- axis of each PC using ImageJ. Each sample was quadruplicated. (i) Overlay of the plot profiles produced in (e-h) highlights well-defined, separated regions of Smo accumulation within the PC. (j) Progression from a proximal based accumulation to a uniform accumulation of Smo throughout the PC correlates with Hh pathway activation.
Figure 7. Key Hh components co-localize with Smo at the ciliary tip. (a) Imaging of EGFP::Gli2 / Smo::tagRFP-T cells shows precise co-localization of Smo and Gli2 at the distal tip of PC.(b) Imaging of Smo::EGFP / SuFu::mCherry cells shows precise co-localization of Smo and SuFu at the distal tip of PC.(a-b) Cells were co-stained for acetylated tubulin (AcTub) and γ-tubulin to demar- cate as well as orient the PC. Scale bar: 5μm.
Figure 7. Key Hh components co-localize with Smo at the ciliary tip. (a)
Imaging of EGFP::Gli2 / Smo::tagRFP-T cells shows precise co-localization of Smo and Gli2 at the distal tip of PC.(b) Imaging of Smo::EGFP / SuFu::mCherry cells shows precise co-localization of Smo and SuFu at the distal tip of PC.(a-b) Cells were co-stained for acetylated tubulin (AcTub) and γ-tubulin to demar- cate as well as orient the PC. Scale bar: 5μm.

Gli2 cannot serve as a viable reporter for Smo-mediated Hh transcriptional activation

A recent study of Gli2 has identified Gli2 trafficking as a potential link between Hh-dependent Smo activation in the PC and transcriptional activation in the nucleus (Kim, 2009). This raised the possibility that Gli2 might serve as a viable reporter for Hh transcriptional status in the nucleus and potentially provide the connection between a particular pattern of Smo distribution at the PC and its functional outcome. To test this possibility, we introduced EGFP::Gli2 and Smo::tagRFP-T into Hh responsive NIH3T3 cells to generate a clonal cell line in which the behavior and function of EGFP::Gli2 mirrored those of endogenous Gli2 (Figure 8a-8c). This then allowed for the simultaneous visualization of Smo and Gli2 within the same cell. However, a closer observation of Smo and Gli2 suggested that each may accumulate at the PC independent of the other. In a small subset of cells, robust Smo accumulation at the PC could be observed without the presence of Gli2, and vice versa (Figure 8d). One explanation for this discrepancy may be that upon the activation of Smo, Gli2 is quickly trafficked out of the PC and into the nucleus to activate the transcription of Hh target genes. The case where Gli2 is present at the PC while Smo is not may simply be due to a snapshot effect of Gli2 cycling in and out of the PC in the absence of Hh stimulation. Because it is impossible to draw any conclusion about the Hh transcriptional status by looking at a fixed image of Gli2, we concluded that Gli2 cannot serve as a viable reporter for Smo-mediated Hh transcriptional activation.

Figure 8. Smo and Gli2 may accumulate at the PC independent of the other. (a) Comparative imaging analysis of EGFP::Gli2 and endogenous Gli2 confirms that EGFP::Gli2 mirrors the movement of endogenous Gli2 under standard treatments. Scale bar: 5μm. (b) Quantification of images produced in (a).(c) Gli-Luciferase reporter assay comparing the transcriptional response in EGFP::Gli2 / Smo::tagRFP-T cells versus NIH3T3 cells. No significant differ- ences in Hh response between the two cell lines could be seen across the stan- dard treatments. Mean (±S.D.) was calculated from six replicate samples.(d) Smo accumulation in the PC does not always predict Gli2 presence in the PC, and vice versa. Cells were immunostained for acetylated tubulin (AcTub) to la- bel the PC. Dotted white lines were drawn to demarcate the entire length of the PC on the basis of overexposed, high-contrast image settings. Scale bar: 5μm.
Figure 8. Smo and Gli2 may accumulate at the PC independent of the other. (a) Comparative imaging analysis of EGFP::Gli2 and endogenous Gli2 confirms that EGFP::Gli2 mirrors the movement of endogenous Gli2 under standard treatments. Scale bar: 5μm. (b) Quantification of images produced in (a).(c) Gli-Luciferase reporter assay comparing the transcriptional response in EGFP::Gli2 / Smo::tagRFP-T cells versus NIH3T3 cells. No significant differ- ences in Hh response between the two cell lines could be seen across the stan- dard treatments. Mean (±S.D.) was calculated from six replicate samples.(d) Smo accumulation in the PC does not always predict Gli2 presence in the PC, and vice versa. Cells were immunostained for acetylated tubulin (AcTub) to la- bel the PC. Dotted white lines were drawn to demarcate the entire length of the PC on the basis of overexposed, high-contrast image settings. Scale bar: 5μm.

Sub-ciliary distribution of Smo is functionally meaningful

To determine the functional relevance of the different patterns of Smo distribution observed at the PC, a more direct assay, Gli1 mRNA FISH, was performed to visualize Gli1 mRNA transcripts in Smo::tagRFP-T cells (Figure 9a-9f). By coupling Gli1 mRNA FISH with Smo::tagRFP-T cells, the position of Smo at the PC and the concurrent Hh transcriptional state could be simultaneously visualized in the same cell. To quantify the levels of Gli1 mRNA transcripts, green dots enclosed by the area of the nucleus were counted for each type of cell expressing a particular pattern of Smo distribution at the PC. This method of counting ensured that any green dot counted belonged exclusively to the cell of interest. The mean area of nuclei across the cells analyzed was more or less equal, demonstrating that any difference seen in the levels of Gli1 mRNA transcripts was not due to a difference in the area of nucleus (Figure 9g). Higher levels of Gli1 mRNA transcripts were associated with dumbbell-shaped or uniform accumulation of Smo relative to proximal or distal based accumulation of Smo. Collectively, these results demonstrate that the position of Smo at any time point along the length of the PC makes a functional difference in the Hh transcriptional outcome.

Figure 9. Smo position within the PC is functionally meaningful. (a-e) Gli1 mRNA FISH assay of Smo::tagRFP-T cells, in which each cell analyzed possessed one of the defined patterns of Smo distribution, including (a) blank, (b) proximal based, (c) distal based, (d) dumbbell-shaped, and (e) uniform accumulation of Smo throughout the PC. Smo occupancy at the ciliary tip is correlated with increased Hh transcriptional activity. Cells were co-stained for acetylated tubulin (AcTub) and γ-tubulin to demarcate as well as orient the PC. White arrow heads indicate the spot of γ-tubulin staining. White dotted lines were drawn to outline the cell nuclei (not shown in the images). Each green dot corresponds to a Gli1 mRNA transcript. Scale bar: 3μm.(f ) Quantification of the images produced in (a-e). For each cell, only the green dots enclosed by the outline of the cell nucleus were counted to ensure ownership by the cell of interest. * indicates statistical significance in a t test comparing to the blank control (***p<0.001).(g) The mean area of nuclei is consistent across the cells possessing any one of the defined patterns of Smo distribution at the PC.
Figure 9. Smo position within the PC is functionally meaningful. (a-e)
Gli1 mRNA FISH assay of Smo::tagRFP-T cells, in which each cell analyzed possessed one of the defined patterns of Smo distribution, including (a) blank, (b) proximal based, (c) distal based, (d) dumbbell-shaped, and (e) uniform accumulation of Smo throughout the PC. Smo occupancy at the ciliary tip is correlated with increased Hh transcriptional activity. Cells were co-stained for acetylated tubulin (AcTub) and γ-tubulin to demarcate as well as orient the PC. White arrow heads indicate the spot of γ-tubulin staining. White dotted lines were drawn to outline the cell nuclei (not shown in the images). Each green dot corresponds to a Gli1 mRNA transcript. Scale bar: 3μm.(f ) Quantification of the images produced in (a-e). For each cell, only the green dots enclosed by the outline of the cell nucleus were counted to ensure ownership by the cell of interest. * indicates statistical significance in a t test comparing to the blank control (***p<0.001).(g) The mean area of nuclei is consistent across the cells possessing any one of the defined patterns of Smo distribution at the PC.

Discussion

Smo continually shuttles in and out of the PC in order to maintain Hh responsiveness

In the absence of Hh stimulation, a small subset of cells had invariably accumulated Smo at the PC. This basal level of Smo accumulation at the PC was associated with a weak Hh transcriptional activity, both of which could be effectively suppressed by a Smo antagonist, SANT-1. Together, the evidence suggests that Smo is continually shuttling in and out of the PC in an unstimulated cell. Because the cycling of Smo is not synchronous across cells, a snapshot of a large population of cells would reasonably show a subset of cells with filled PC. This finding is also consistent with a previously study reporting that RNAi knockdown of dynein 2, a retrograde ciliary motor protein, leads to the constitutive ciliary accumulation of Smo (Kim, 2009). The biological advantage in having Smo continually shuttle in and out of the PC may be two-fold. First, it maintains Hh responsiveness in unstimulated cells; Smo translocation to the PC is a requirement for Hh pathway activation, and maintaining low levels of Smo at the PC is essential for the initiation of Hh signaling. Secondly, having Smo already present at the PC where key molecular interactions for Hh activation take place enables a rapid and robust firing of Hh pathway when a ligand activates the pathway. This mechanism may potentially underlie the surprisingly rapid kinetics of Smo accumulation at the PC occurring within minutes after Hh stimulation.

Accumulation of Smo at the ciliary tip is essential for the relay of Hh signaling

A close observation of Smo at the PC revealed an array of discrete patterns of Smo distribution at the PC, including [a] proximal based accumulation, [b] distal based accumulation, [c] dumbbell-shaped accumulation, and [d] uniform accumulation of Smo filling up the entire PC. Furthermore, the frequency at which each pattern was observed shifted over the course of Hh signaling such that a proximal based accumulation of Smo dominated at earlier time points but gradually progressed to patterns of Smo accumulation that commonly occupied the distal tip of PC. Consistent with the Smo imaging data, a transcriptional assay showed higher expression levels of Gli1 mRNA transcripts in cells that had Smo occupying the ciliary tip. Collectively, these results suggest a strong correlation between Smo accumulation at the ciliary tip and its active status. One possible model is that Smo first accumulates at the proximal end of the PC, which functions as the primary docking site for Smo. Once it is relieved from Ptch1 inhibition by a ligand, Smo becomes activated and is then able to translocate to the tip of the PC, resulting in a dumbbell-shaped pattern of Smo distribution. Finally, as increasing amounts of Smo continue to traffic into the PC, Smo eventually fills up the PC, achieving a uniform distribution throughout the entire structure. This model is also consistent with the observation that key components of the Hh pathway all selectively localize at the distal tip of PC, including Gli1-3 and SuFu. Therefore, it is reasonable to infer that Smo accumulation at the ciliary tip enables its interaction with the downstream components of the pathway for the relay of Hh signaling. Although unclear as to how this interaction is achieved, it presumably leads to altered Gli trafficking and processing.

Primary cilum: more complex than previously thought

Restricted patterns of Smo distribution at the PC suggest that the PC is truly compartmentalized, consisting of biologically distinct domains whose entry and exit is tightly regulated. In fact, since the advance of electron microscopy from the 1950s to the 1970s, a careful and systematic ultra-structural analysis of PC has generated an anatomical map of the PC in which the cilium can be dissected into distinct substructural zones, including: the basal body, the transition zone, the doublet zone, and the singlet zone with the ciliary tip complex (see Czarnecki and Shah, 2012). Of these, the transition zone has been linked to a set of remarkable protein assemblies that function as a diffusion barrier and a gate-keeper, maintaining the PC as a privileged membrane domain; a septin-containing diffusion barrier was identified near the base of PC (Chih, 2011; Hu, 2010).

Smo, a 7-pass transmembrane protein, translocates to the PC from an intracellular store via directed vesicular trafficking from the Golgi apparatus (Pazour, 2008; Wang, 2009). Membrane vesicles containing cargos targeted for ciliary transport such as Smo must undergo fusion at the transition zone of the PC (Rosenbaum, 2002). Hence, the transition zone serves as a docking site that potentially underlies a rate limiting step, consistent with the observation of a proximal based accumulation of Smo occurring initially in time. Once it crosses the septin diffusion barrier, Smo is actively transported to the ciliary tip. In this model, the cytoplasmic tail of Smo remains attached to an anterograde motor protein, restricting the random diffusion of Smo. Upon its arrival at the ciliary tip, Smo is released from its trafficking machinery. Subsequently, the coupling of Smo with a separate retrograde motor protein may take place in order to transport Smo out of the PC. This switch from anterograde to retrograde trafficking machinery potentially underlies a second rate limiting step, consistent with the observation of a distal based or a dumbbell-shaped pattern of ciliary accumulation. As large amounts of Smo are rapidly trafficked into the PC upon Hh stimulation, the release of Smo at the ciliary tip may overwhelm the retrograde trafficking machinery, resulting in a uniform accumulation of Smo at the PC.

This model accounts for two independent observations of Smo reported in the literature: [1] rapid fluorescence recovery of Smo::YFP after photobleaching a distal portion of the PC, and [2] ligand-independent accumulation of Smo at the PC after depletion of dynein 2, a retrograde motor protein (Hu, 2010; Kim, 2009). The release of Smo at the ciliary tip creates a pool of Smo that is free to diffuse throughout the ciliary membrane, accounting for its fluorescence recovery after photobleaching. However, fluorescence did not recover when the entire PC was photobleached (Hu, 2010). Hence, Smo is likely trapped within the ciliary membrane by the septin diffusion barrier at the base of the PC, requiring dynein 2 for its ciliary exit.

Materials and Methods

Constructs and Cell Lines

Smo::EGFP, Smo::tagRFP-T, EGFP::Gli2, and Ivs:: tagRFPT were cloned into pBabe retroviral constructs. Subsequently, Smo::EGFP / Ivs::tagRFP-T, Smo::tagRFP-T, EGFP::Gli2/Smo::tagRFPT, and Ivs::tagRFP-T monoclonal, stable cell lines were generated by virally infecting NIH3T3 cells. Clones for each cell line were analyzed by imaging and Gli-Luciferase reporter assay; clones that showed low level expression and expected ciliary trafficking behaviors for each protein of interest were selected for further study. Smo::EGFP / SuFu::mCherry cell line was previously established by Y. Wang (Harvard University, Cambridge, MA) via similar methods.

Cell Culture

NIH3T3 cells and their derivatives were maintained in DMEM containing 10% (v/v) calf serum, penicillin, streptomycin, and L-glutamine. HEK293 and cos7 cells were maintained in DMEM containing 10% (v/v) fetal bovine serum, penicillin, streptomycin, and L-glutamine. Shh conditioned medium was generated by transfecting cos7 cells with an expression construct encoding the amino terminal 19kDa signaling peptide of Shh. Conditioned medium was harvested 3-4 days after the cells reached confluency. Control conditioned medium was collected from cos7 cells transfected with an empty plasmid.

Imaging Assays

Cells were cultured and treated in 384-well imaging plate pre-coated with poly-D-Lysine (Greiner Bio-one). Images were collected using Opera High Content Screening System (Perkin Elmer). Acapella 2.0 software (Evotec Technologies/PerkinElmer) was used to perform multi-parametric image quantification. All the comparative images were scanned with identical microscopic setting and analyzed with the same input parameters. Excel (Microsoft), Prism (GraphPad), ImageJ (NIH), and Photoshop (Adobe) were used for data analysis and/or editing.

Luciferase Assays

Cells were plated at a density of 1×104 per well in 96-well plates (Corning) 18 hours before transfection. DNA transfection was performed using FuGene HD (Roche); the DNA introduced in each well included 25ng of CMV driven renilla Luciferase construct and 75ng of Ptch1 promoter driven firefly Luciferase construct. Cells were cultured for 2 days prior to treatments and for another 2 days prior to the assay. Cells were subsequently assayed using a Promega dual Luciferase reporter assay system kit (E1910). Renilla Luciferase signal was used to normalize the firefly Luciferase signal. Luciferase signal was read by TopCount NX Microplate Scintillation and Luminescence Counter (Perkin Elmer). In all Luciferase assays, plates were read three consecutive times to produce average measurements.

Fluorescence RNA in situ Hybridization (FISH) Assays

Cells were cultured and treated in 384-well imaging plate pre-coated with poly-D-Lysine (Greiner Bio-one). After treatment, cells were fixed and processed according to the protocol for fluorescence RNA in situ hybridization (FISH) provided by Panomics. All reagents used in the assay were purchased from Panomics, including the RNA target probe sets for Gli1 and GAPDH. Following FISH, cells were antibody stained for acetylated tubulin and γ-tubulin for the visualization and orientation of primary cilia. Plates were finally imaged using the Opera High Content Screening System, and all subsequent image quantifications were produced by a modified script using Acapella 2.0 software.

References

Bai CB, Auerbach W, Lee JS, Stephen D, Joyner AL. (2002). Gli2, but not Gli1, is required for initial Shh signaling and ectopic activation of the Shh pathway. Development 129:4753–4761.

Barakat MT, Humke EW, Scott MP. (2010). Learning from Jekyll to control Hyde: Hedgehog signaling in development and cancer. Trends Mol Med 16(8):337-48.

Chih B, Liu P, Chinn Y, Chalouni C, Komuves LG, Hass PE, Sandoval W, Peterson AS. (2011). A ciliopathy complex at the transition zone protects the cilia as a privileged membrane domain. Nat Cell Biol 14(1):61-72.

Corbit KC, Aanstad P, Singla V, Norman AR, Stainier DY, et al. (2005). Vertebrate Smoothened functions at the primary cilium. Nature 437:1018–1021.

Czarnecki PG, Shah JV. (2012). The ciliary transition zone: from morphology and molecules to medicine. Trends Cell Biol [Epub ahead of print].

Ding Q, Fukami S, Meng X, Nishizaki Y, Zhang X, et al. (1999). Mouse suppressor of fused is a negative regulator of sonic hedgehog signaling and alters the subcellular distribution of Gli1. Curr Biol 9:1119–1122.

Eggenschwiler JT, Anderson KV. (2007). Cilia and developmental signaling. Annu Rev Cell Dev Biol 23:345–373.

Goetz, S.C. & Anderson, K.V. (2010). The primary cilium: a signalling centre during vertebrate development. Nat Rev Genet 11:331-44.

Haycraft CJ, et al. (2005). Gli2 and Gli3 localize to cilia and require the intraflagellar transport protein polaris for processing and function. PLoS Genet 1:e53.

Hooper JE, Scott MP. (2005). Communicating with Hedgehogs. Nat Rev Mol Cell Biol 6:306-317.

Hu Q, Milenkovic L, Jin H, Scott MP, Nachury MV, Spiliotis ET, Nelson WJ. (2010). A septin diffusion barrier at the base of the primary cilium maintains ciliary membrane protein distribution. Science 329(5990):436-9.

Huangfu D, Anderson KV. (2005). Cilia and Hedgehog responsiveness in the mouse. Proc Natl Acad Sci U.S.A. 102:11325–11330.

Huangfu D, et al. (2003). Hedgehog signalling in the mouse requires intraflagellar transport proteins. Nature 426:83–87.

Ingham PW, McMahon AP. (2001). Hedgehog signaling in animal development: Paradigms and principles. Genes Dev 15:3059–3087.

Jiang J, Hui CC. (2008). Hedgehog signaling in development and cancer. Dev Cell 15:801–812.

Kim J, Kato M, Beachy PA. (2009). Gli2 trafficking links Hedgehog-dependent activation of Smoothened in the primary cilium to transcriptional activation in the nucleus. Proc Natl Acad Sci U.S.A. 106(51):21666-71.

Kiprilov EN, et al. (2008). Human embryonic stem cells in culture possess primary cilia with hedgehog signaling machinery. J Cell Biol 180:897–904.

Kogerman P, Grimm T, Kogerman L, Krause D, Unden AB, et al. (1999). Mammalian suppressor-of-fused modulates nuclear-cytoplasmic shuttling of Gli1. Nat Cell Biol 1:312–319.

McMahon AP, Ingham PW, Tabin CJ. (2003). Developmental roles and clinical significance of hedgehog signaling. Curr Top Dev Biol 53:1–114.

Methot N, Basler K. (2000). Suppressor of fused opposes hedgehog signal transduction by impeding nuclear accumulation of the activator form of Cubitus interruptus. Development 127:4001–4010.

Park HL, et al. (2000). Mouse Gli1 mutants are viable but have defects in SHH signaling in combination with a Gli2 mutation. Development 127:1593–1605.

Pazour, G.J., R.A. Bloodgood. (2008). Targeting proteins to the ciliary membrane. Curr Top Dev Biol 85:115–149.

Pearse RV, 2nd, Collier LS, Scott MP, Tabin CJ. (1999). Vertebrate homologs of Drosophila suppressor of fused interact with the Gli family of transcriptional regulators. Dev Biol 212:323–336.

Persson M, et al. (2002). Dorsal-ventral patterning of the spinal cord requires Gli3 transcriptional repressor activity. Genes Dev 16:2865–2878.

Pugacheva EN, Jablonski SA, Hartman TR, Henske EP, Golemis EA. (2007). HEF1- dependent Aurora A activation induces disassembly of the primary cilium. Cell 129:1351–1363.

Rohatgi R, Milenkovic L, Scott MP. (2007). Patched1 regulates hedgehog signaling at the primary cilium. Science 317:372–376.

Romer, J. T. et al. (2004). Suppression of the Shh pathway using a small molecule eliminates medulloblastoma in Ptc1(+/-)p53(-/-) mice. Cancer Cell 6:229-240.

Rosenbaum, J.L., G.B. Witman. (2002). Intraflagellar transport. Nat Rev Mol Cell Biol 3:813–825.

Tran PV, et al. (2008). THM1 negatively modulates mouse sonic hedgehog signal transduction and affects retrograde intraflagellar transport in cilia. Nat Genet 40:403–410.

Tukachinsky H, Lopez LV, Salic A. (2010). A mechanism for vertebrate Hedgehog signaling: recruitment to cilia and dissociation of SuFu-Gli protein complexes. J Cell Biol 191(2):415-28.

Wang B, Fallon JF, Beachy PA. (2000). Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate limb. Cell 100:423–434.

Wang Y, Zhou Z, Walsh CT, McMahon AP. (2009). Selective translocation of intracellular Smoothened to the primary cilium in response to Hedgehog pathway modulation. Proc Natl Acad Sci U.S.A. 106(8):2623-8.

Watanabe D, et al. (2003). The left-right determinant Inversin is a component of node monocilia and other 9+0 cilia. Development 130:1725-1734.

Wilson CW, Chen MH, Chuang PT. (2009). Smoothened adopts multiple active and inactive conformations capable of trafficking to the primary cilium. PLoS One 4(4):e5182.

Xie, J. et al. (1998). Activating Smoothened mutations in sporadic basal-cell carcinoma. Nature 391:90-92

Zeng H, Jia J, Liu A. (2010). Coordinated translocation of mammalian Gli proteins and suppressor of fused to the primary cilium. PLoS One 5(12):e15900.

Comments:

NO COMMENTS

LEAVE A REPLY