Cover ArticleResearch Articles, Development/Plasticity/Repair

β-arrestins Are Scaffolding Proteins Required for Shh-Mediated Axon Guidance

Rachelle Sauvé, Steves Morin, Patricia T. Yam and Frédéric Charron
Journal of Neuroscience 24 July 2024, 44 (30) e0261242024; https://doi.org/10.1523/JNEUROSCI.0261-24.2024

Abstract

During nervous system development, Sonic hedgehog (Shh) guides developing commissural axons toward the floor plate of the spinal cord. To guide axons, Shh binds to its receptor Boc and activates downstream effectors such as Smoothened (Smo) and Src family kinases (SFKs). SFK activation requires Smo activity and is also required for Shh-mediated axon guidance. Here we report that β-arrestin1 and β-arrestin2 (β-arrestins) serve as scaffolding proteins that link Smo and SFKs in Shh-mediated axon guidance. We found that β-arrestins are expressed in rat commissural neurons. We also found that Smo, β-arrestins, and SFKs form a tripartite complex, with the complex formation dependent on β-arrestins. β-arrestin knockdown blocked the Shh-mediated increase in Src phosphorylation, demonstrating that β-arrestins are required to activate Src kinase downstream of Shh. β-arrestin knockdown also led to the loss of Shh-mediated attraction of rat commissural axons in axon turning assays. Expression of two different dominant-negative β-arrestins, β-arrestin1 V53D which blocks the internalization of Smo and β-arrestin1 P91G-P121E which blocks its interaction with SFKs, also led to the loss of Shh-mediated attraction of commissural axons. In vivo, the expression of these dominant-negative β-arrestins caused defects in commissural axon guidance in the spinal cord of chick embryos of mixed sexes. Thus we show that β-arrestins are essential scaffolding proteins that connect Smo to SFKs and are required for Shh-mediated axon guidance.

Significance Statement

The correct guidance of axons is important for the formation of the nervous system. Sonic hedgehog (Shh)-mediated axon guidance relies on the activation of Src family kinases (SFKs) downstream of the atypical G-protein-coupled receptor (GPCR) Smoothened (Smo). How SFKs are activated downstream of Smo was unknown. In this study, we found that β-arrestin1 and β-arrestin2 (β-arrestins) serve as scaffolding proteins between Smo and SFKs. We also found that β-arrestins are required for the activation of SFKs. Knocking down β-arrestins or expressing dominant-negative β-arrestins caused loss of Shh-mediated attraction of commissural axons. In vivo, the expression of dominant-negative β-arrestins caused commissural axon guidance defects. Our work identifies for the first time a role for β-arrestins in axon guidance.

Introduction

During nervous system development, Sonic hedgehog (Shh) acts as a morphogen to induce cell fate specification and also acts as an axon guidance cue to direct axons to their targets (Yam and Charron, 2020). Cell fate specification by Shh occurs via the canonical Shh signaling pathway, where binding of Shh to its receptor, Patched (Ptch1) together with the coreceptors Boc, Cdon and Gas1, relieves inhibition of the transmembrane protein Smoothened (Smo), a primary activator of the pathway. Smo activation leads to downstream signaling events which ultimately results in Gli-dependent transcription (Ingham and McMahon, 2001; Allen et al., 2011; Izzi et al., 2011). Shh is also an attractive axon guidance cue for spinal cord commissural neurons. In the developing spinal cord, Shh, secreted by floor plate cells, attracts axons to the ventral midline (Charron et al., 2003). Shh guides axons through a noncanonical Shh signaling pathway, where Shh binds to Boc and Ptch1. Shh binding relieves the Ptch1 repression of Smo leading to Smo activation and subsequent activation of Src family kinases (SFKs; Okada et al., 2006; Yam et al., 2009). Active SFKs phosphorylate Zipcode binding protein 1 (ZBP1), mediating its release from β-actin mRNA which is then locally translated for axon attraction (Lepelletier et al., 2017). However, the process by which SFKs are activated downstream of Smo is unknown.

The arrestin protein family is composed of two visual arrestins, arrestin 1 and arrestin 4, that are specific to rods and cones, and two nonvisual arrestins, named β-arrestin1 and β-arrestin2 (β-arrestins, or arrestin 2 and 3). β-arrestins are known for their role in G-protein-coupled receptor (GPCR) desensitization, endocytosis, and signaling (Lefkowitz and Shenoy, 2005). Upon GPCR stimulation, β-arrestins are recruited to the GPCR following phosphorylation of the cytoplasmic tail of the receptor by G-protein-coupled receptor kinases (GRKs). β-arrestin binding to the activated GPCR can cause receptor desensitization, in which G-protein coupling to the receptor is impaired and G-protein signaling is decreased (Ferguson, 2001). β-arrestins also participate in clathrin-mediated endocytosis of activated GPCRs, an essential process for receptor recycling or degradation (DeWire et al., 2007).

Smo is a seven span transmembrane protein with a similar sequence and protein architecture as Frizzled receptors, making it a member of the class F superfamily of GPCRs (Alexander et al., 2017). In canonical Shh signaling, β-arrestins bind to Smo and are required for Smo localization to the primary cilium, where Smo is activated (Kovacs et al., 2008). β-arrestins also promote the endocytosis of Smo via clathrin-coated pits (W. Chen et al., 2004). β-arrestins are required for Shh-induced Gli-dependent transcription (Kovacs et al., 2008), and in zebrafish, β-arrestin2 knockdown causes many defects characteristic of impaired canonical Shh signaling, including ventrally curved bodies and U-shaped somites, a phenotype similar to Smo or Gli2 mutants (W. Chen et al., 2001; Karlstrom et al., 2003; Wilbanks et al., 2004).

β-arrestins can also act as scaffolding proteins, recruiting signaling proteins such as SFKs to GPCRs to mediate downstream signaling (Peterson and Luttrell, 2017). For example, stimulation of the β2 adrenergic receptor (β2AR) increases the association between β-arrestins and Src, and this is required for the activation of the Erk signaling pathway downstream of β2AR (Luttrell et al., 1999). Unlike canonical Shh signaling, which does not require SFK activity, noncanonical Shh signaling in axon guidance requires SFKs (Yam et al., 2009). Given that β-arrestins can interact with Smo and also with SFKs, we hypothesized that a novel role for β-arrestins could be to act as scaffolding proteins linking Smo to SFK activation in Shh-mediated commissural axon guidance.

Here, we found that β-arrestins recruit SFKs to Smo to form a tripartite complex. We also found that β-arrestins are required for the activation of SFKs, a process that is specific to noncanonical but not canonical Shh signaling. We then demonstrate that β-arrestins are required for Shh-mediated commissural axon guidance. Altogether, our findings identify β-arrestins as essential scaffold proteins for Shh-mediated axon guidance.

Materials and Methods

Animals

All animal work was performed in accordance with the Canadian Council on Animal Care Guidelines and approved by the Montreal Clinical Research Institute Animal Care Committee. Staged pregnant female Sprague Dawley rats were obtained from Charles River Laboratories (Kingston). Embryonic day 0 (E0) was defined as midnight of the night before a plug was found. Tissue from embryos of either sex was used for experiments.

Primary commissural neuron culture

Commissural neurons from Sprague Dawley rat embryos at E13.5 were dissected as described previously (Langlois et al., 2010). Briefly, commissural neurons from the dorsal fifth region of the spinal cord were isolated. Commissural neurons were plated in Neurobasal (Life Technologies, 21103049) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Wisent) and 2 mM GlutaMAX (Life Technologies, 35050061). The medium was changed to Neurobasal supplemented with 2% B27 (Life Technologies, 17504044) and 2 mM GlutaMAX after ∼20 h. Commissural neurons were maintained at 5% CO2 in a humidified incubator, and they were used for experiments after 2 d of in vitro culture (Langlois et al., 2010). This procedure reliably yields cultures of commissural neurons with a high degree of purity, with ∼90% of neurons expressing the commissural neuron marker DCC and almost 100% of neurons expressing the commissural neuron marker LH2 (Yam et al., 2009). Each experiment was performed using cells pooled from the same litter. For Dunn chamber experiments, electroporated commissural neurons were plated at 210,000–330,000 cells/well in a six-well plate on acid-washed 18 mm square #3D coverslips (Assistent) coated with poly-ʟ-lysine (PLL, Sigma, P4707). For biochemical experiments, commissural neurons were plated at >500,000 cells/well in a six-well plate or at >1.0 × 106 cells in a 60 mm tissue culture dish (Corning) coated with PLL. When needed, cells were treated for 30 min with a bath application of 0.1 µg/ml recombinant Shh (R&D Systems, 8908-SH) or BSA (Sigma, A4161), the vehicle for Shh, prior to harvesting.

Commissural neuron electroporation

Commissural neurons were electroporated with the Amaxa 96-well Shuttle using the P3 Primary Cell Nucleofector Kit (Lonza). For each electroporation in one well of a 96-well Nucleofector plate, 3.0–7.5 × 105 commissural neurons were electroporated with 0.4 µg of plasmid DNA or with 2 µM nontargeting scrambled siRNA or a mix of siRNA targeting β-arrestin1/2 (1 µM β-arrestin1 siRNA + 1 µM β-arrestin2 siRNA) according to the manufacturer's instructions with the program DC-100 or CP-100. Greater than 80% of commissural neurons successfully expressed the plasmid after electroporation.

Small interfering RNA

siRNAs were designed using the Custom Dicer-Substrate siRNA (DsiRNA) system (IDT). siRNA oligonucleotides were reconstituted in H2O, aliquoted, and stored at −20°C.

Scrambled:

5′-CUUCCUCUCUUUCUCUCCCUUGUGA-3′

3′-AGGAAGGAGAGAAAGAGAGGGAACACU-5′

β-arrestin1 siRNA:

5′-GGUAUCAUUGUUUCCUACAAAGUCA-3′

3′-ACCCAUAGUAACAAAGGAUGUUUCAGU-5′

β-arrestin2 siRNA:

5′-CUCAUUGAAUUCGAUACCAACUATG-3′

3′-UGGAGUAACUUAAGCUAUGGUUGAUAC-5′

Plasmids

pcDNA3-β-arrestin1-WT-Myc was a gift from M. Bouvier. pcDNA3-β-arrestin1-V53D-Myc and pcDNA3-β-arrestin1-P91G-P121E-Myc were generated by mutating β-arrestin1-WT-Myc using the QuikChange Multi Site-Directed Mutagenesis Kit (Stratagene). β-Arrestin1-WT-Myc, β-Arrestin1-V53D-Myc and β-arrestin1-P91G-P121E-Myc coding sequences were subcloned into the pCAGGS vector. Sequences were verified by Sanger sequencing. Smo-GFP was subcloned from the pEGFP-N1 vector into pCAGGS vector (Izzi et al., 2011). WT-ShhN was in the pMT21 vector (Izzi et al., 2011). The GFP construct used was pCAGGS-EGFP and Fyn-GST in the pDEST27 vector was a gift from R. Screaton. J2X-GFP was used to express GFP under Math1 promoter (Lepelletier et al., 2017).

Cell lines

Cos7 and NIH3T3 (Gli-Luc) cells were maintained in DMEM (Life Technologies, 11995065) + 10% FBS + penicillin/streptomycin (Life Technologies, 15140122) in a 5% CO2 incubator. When needed, cells were serum starved 24 h after transfection and treated 48 h after transfection. Cos7 cells were treated with 20 nM Shh (R&D Systems, 1845-SH) for 1 h.

ShhN conditioned media production

Cos7 cells were transfected using Lipofectamine 3000 (Life Technologies, L3000015) with pMT21 (empty vector) or pMT21-WT-ShhN. Cells were then cultured in Opti-MEM (Life Technologies, 31985070) without serum and the media was collected 1–2 d after transfection. The amount of Shh present in the conditioned media was evaluated by dot blot on nitrocellulose membrane by loading different volumes of the media and comparing these signals to a standard curve generated using 20–1,000 ng ShhN (R&D Systems, 1845-SH).

Coimmunoprecipitation

Commissural neurons were lysed on ice in SLB lysis buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% Igepal) complemented with protease (Roche, 11873580001) and phosphatase inhibitors (Roche, 4906837001). Protein concentration was determined with the BCA assay (Thermo Fisher Scientific, 23227). Cos7 cells were transfected with plasmids using Lipofectamine 2000 (Life Technologies, 11668019) and lysed 48 h after transfection in RIPA lysis buffer (50 mM HEPES, pH 7.4, 150 mM NaCl, 10% glycerol, 1.5 mM MgCl2, 1 mM EDTA, 1% Triton X-100, 0.1% SDS) complemented with protease and phosphatase inhibitors. The lysate was precleared 30 min with Protein A/G PLUS-Agarose beads (Santa Cruz, sc-2003). Equal amounts of protein lysate were incubated with beads and with the indicated antibodies for 4 h at 4°C on a rotator. The following antibodies were used for the immunoprecipitation: anti-GFP (Molecular Probes, A11120), anti-Src (A. Veillette) or a mouse IgG isotype control (Abcam, ab37355). The immunoprecipitated protein complexes were pulled down at 2,000 rpm at 4°C for 2 min and beads were washed three times with lysis buffer before adding sample buffer and boiled for 5 min at 95°C. The immunoprecipitated proteins were analyzed by Western blotting.

GST pull down

Cos7 cells were transfected using Lipofectamine 2000 and lysed in RIPA buffer complemented with protease and phosphatase inhibitors. Fyn-GST was pulled down (PD) using glutathione beads. PD proteins were analyzed by Western blotting.

Western blotting

After boiling in sample buffer, equal amount of samples were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were transferred onto polyvinylidene difluoride membrane (PVDF), and membranes were blocked with 5% BSA in TBST (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20) or with 5% skim milk in TBST. Membranes were incubated with primary antibodies in 5% BSA in TBST + 0.02% sodium azide followed by an incubation with secondary antibodies conjugated to horseradish peroxidase (HRP). The following primary antibodies were used: anti-Myc (Santa Cruz Biotechnology, sc-40), anti-GST (Santa Cruz Biotechnology, sc-138), anti-GFP (Molecular Probes, A11121), anti-Smo (Santa Cruz Biotechnology, sc-166685), anti-Src (Millipore, 05-184), anti-pan-arrestin (Abcam, ab2914), anti-pSrc Y418 (Invitrogen, 44660G) and anti-GAPDH (Millipore, MAB374). The secondary antibodies used were anti-mouse IgG-HRP (Jackson ImmunoResearch, 115-035-003) and anti-rabbit IgG-HRP (Jackson ImmunoResearch, 111-035-003). Protein bands were visualized with chemiluminescence.

In situ hybridization

In situ mRNA detection of E11.5 mouse embryo spinal cord cross sections was performed as previously described (Schaeren-Wiemers and Gerfin-Moser, 1993; Kania and Jessell, 2003). Using a mouse cDNA library, we generated probes targeting β-arrestin1 and β-arrestin2. The size of each probe was 700–1,000 bp. The primer sequences used to make the probes are as follows:

  • Arrb1 (forward): 5′ ACA CAA GAA GAT GTC TGT AG 3′

  • Arrb1 (reverse): 5′ GAC TCT GGG TAC TAA TAA AA 3′

  • Arrb2 (forward): 5′ TAT CAT CAG AAA GGT ACA GT 3′

  • Arrb2 (reverse): 5′ TTT CTA CAC CTT TTC TCT AC 3′

Dunn chamber axon guidance assay and analysis

This assay was performed as described previously (Yam et al., 2009). Briefly, commissural neurons were dissociated and electroporated as described above and plated on PLL-coated #3D square coverslips (Assistent). After 2 d of culture, commissural neurons were exposed to a gradient of Shh conditioned media or the control empty vector conditioned media at a concentration of 0.1–0.4 µg/ml in the outer well of the Dunn chambers. Commissural neurons were monitored by time-lapse phase contrast microscopy every 4 min over 2 h at 37°C with a 10× fluotar or 20× fluotar objective on a Leica DMIRE2 inverted microscope (Leica) equipped with a MS-2000 XYZ automated stage (Applied Scientific Instrumentation). The images were acquired on an Orca-ER CCD camera (Hamamatsu) using Volocity (Improvision). The axon trajectories were analyzed with FIJI (National Institutes of Health). The turned angle of the commissural axons was defined as the angle between the original direction of the axon and a straight line connecting the base of the growth cone from the first to the last time point of the assay period.

Chick in ovo electroporation

Chick spinal cord electroporation of plasmids was performed at Hamburger Hamilton (HH) stage 11–12, generally as described previously (Luria et al., 2008). Briefly, plasmid DNA solution at 1–10 mg/ml in TE buffer, pH 7.5 (10 mM Tris-HCl) and 1 mM EDTA was injected into the lumbar neural tube through a small eggshell window. The following plasmids were injected: J2X-GFP with pcDNA3, pCAGGS-β-arrestin1-WT-Myc, β-arrestin1-V53D-Myc, or β-arrestin1-P91G-P121E-Myc. The lower body of the embryos was electroporated using platinum/iridium electrodes (FHC) with the ECM 830 Electro Square Porator (BTX; Harvard Apparatus; INTRAcel; 30 V, 5 pulses 50 ms wide at 1 s interval). The eggshell windows were closed with Parafilm and incubated at 38°C until harvesting at HH 28–29. Embryos were harvested, fixed in 4% PFA, and embedded in OCT (Tissue-Tek OCT, Thermo Fisher Scientific, 1437365). Spinal cords were sectioned using a cryostat and immunostained. The images were analyzed with ImageJ (National Institutes of Health) to measure the area of the GFP+ commissural axon tract in the ventral half of the electroporated side of the neural tube relative to the total area of the ventral half of the electroporated spinal cord.

Immunostaining

NIH3T3 cells were transfected with the indicated plasmids using Lipofectamine 3000, and cells were fixed 48 h after transfection for 15 min by adding an equal volume of 8% PFA (final concentration of 4% PFA) to the cells. Immunostaining of cells and tissue sections were performed by permeabilizing the cells with 0.3% Triton X-100 for 15 min and by blocking the samples with 10% donkey serum and 0.1% Triton X-100 in PBS, pH 7.4, for 1 h at room temperature. The samples were incubated overnight at 4°C with the primary antibody diluted in PBS with 1% donkey serum and 0.1% Triton X-100. The following primary antibodies were used: anti-GFP (Invitrogen, A11122), anti-GFP (Molecular Probes, A11120), anti-Myc (Santa Cruz Biotechnology, sc-40), anti-Myc (Abcam, ab9106), or anti-cleaved caspase-3 (Cell Signaling, 9661). The samples were incubated with the secondary antibody diluted in PBS with 1% donkey serum and 0.1% Triton X-100 for 1 h at room temperature. The following secondary antibodies (Jackson ImmunoResearch) were used: donkey anti-rabbit IgG Alexa Fluor 488 (711-545-152), donkey anti-rabbit IgG Cy3 (711-165-152), donkey anti-rabbit IgG Alex Fluor 647 (711-605-152), goat anti-mouse IgG Alexa Fluor 488 (115-545-166), goat anti-mouse IgG Cy3 (115-165-146), goat anti-rabbit IgG Alexa Fluor 488 (111-545-144), or goat anti-rabbit IgG Cy3 (111-165-144). Nuclei were stained with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich, D95964), and samples were mounted with Mowiol (Sigma-Aldrich, 81381).

Image acquisition

NIH3T3 cells were imaged with a Leica DM6 microscope with a 63× oil objective. Chick embryo spinal cords and bright-field images of in situ hybridization samples were imaged with a Leica DM6000 microscope with an Orca-ER CCD camera (Hamamatsu) using Volocity.

Experimental design and statistical analysis

All statistical analyses were performed using GraphPad Prism version 7.03 for Windows. All data are presented as mean ± SEM. Two-tailed Student’s t tests were used to determine the differences between two groups and one-way ANOVA analysis were used to determine the difference between more than two groups, using Tukey's multiple-comparisons test as a post hoc test. The number of experiments presented (n) in each figure represent independent experiments or the number of animals used.

Results

β-arrestins are expressed in commissural neurons during development

Shh guides commissural axons of the developing spinal cord. To examine if β-arrestins are expressed in spinal cord commissural neurons during development, we analyzed our previously published RNA sequencing data from dissociated commissural neurons isolated from the spinal cord of E13.5 rats (GSE268644; Makihara et al., 2018), a stage where commissural axons are being guided by Shh. We found that both β-arrestin1 and β-arrestin2 transcripts are relatively abundant in cultured embryonic commissural neurons when compared with other members of the noncanonical Shh pathway and that both transcripts are expressed at similar levels (Fig. 1A). We then assessed the localization of β-arrestin1/2 mRNA in the developing mouse embryonic spinal cord at E11.5 (which corresponds to rat E13.5) by in situ hybridization. We observed β-arrestin1 mRNA expression in the commissural neuron region, which was marked by high DCC expression, and in other parts of the spinal cord, except in the ventricular zone. We found that β-arrestin2 mRNA was expressed throughout the spinal cord, including the commissural neuron region (Fig. 1B). Our results show that β-arrestins mRNA are present in developing commissural neurons, indicating that β-arrestins are expressed in embryonic commissural neurons at a stage when they are being guided by Shh.

Figure 1.
Figure 1.

β-arrestins are expressed in commissural neurons during development. A, Fragments per kilobase per million mapped fragments (FPKM) values of β-arrestins mRNA (Arrb1 and Arrb2) and other Shh signaling pathway components from RNA sequencing of dissociated cultured embryonic rat commissural neurons [GSE268644; from Makihara et al. (2018)]. B, In situ hybridization showing β-arrestin1 and β-arrestin2 mRNA expression in E11.5 mouse embryonic spinal cord. DCC expression was used to define the commissural neuron (CN) population. β-arrestin1 and β-arrestin2 mRNA are expressed by commissural neurons in the developing spinal cord. β-arrestin2 mRNA is also expressed by cells in the ventricular zone (VZ). Scale bar, 200 µm.

SFKs interact with Smo via β-arrestins

SFKs, more particularly Src and Fyn, are required for Shh-mediated axon guidance (Yam et al., 2009). To test if β-arrestins could act as scaffolding proteins between Smo and SFKs, we first tested the interaction between β-arrestins and Fyn by doing pull-down assays in Cos7 cells, a cell line expressing low levels of endogenous β-arrestins (Ménard et al., 1997). We coexpressed β-arrestin1-Myc or β-arrestin2-Myc together with Fyn-GST or Fyn-GST alone, and we pulled down Fyn-GST. Both β-arrestins were pulled down with Fyn-GST, indicating that β-arrestins interact with Fyn (Fig. 2A). No Myc signal was detected in the pull-down in the conditions where Fyn-GST, β-arrestin1-Myc, or β-arrestin2-Myc were expressed alone, indicating that there was no nonspecific binding in the assay. This result is consistent with previous reports showing interactions between β-arrestins and SFKs, including Fyn (Luttrell et al., 1999; Barlic et al., 2000; F. Yang et al., 2018).

Figure 2.
Figure 2.

SFKs interact with Smo via β-arrestins. A, Cos7 cells were transfected with Fyn-GST, β-arrestin1-Myc (β-arr1-Myc), and β-arrestin2-Myc (β-arr2-Myc) as indicated. Fyn-GST was pulled down (PD) from the cell lysates using glutathione beads and β-arrestin1-Myc or β-arrestin2-Myc were detected by immunoblotting (IB) with an anti-Myc antibody. Both β-arrestin1-Myc and β-arrestin2-Myc were pulled down with Fyn-GST. B, Smo-GFP was coexpressed with β-arrestin1-Myc or β-arrestin2-Myc in Cos7 cells. Smo-GFP was immunoprecipitated (IP) with an anti-GFP antibody and β-arrestin1-Myc and β-arrestin2-Myc were detected with an anti-Myc antibody. β-arrestin1 and β-arrestin2 coimmunoprecipitated with Smo-GFP. C, Smo-GFP was coexpressed with β-arrestin1-Myc in Cos7 cells and cells were stimulated with 20 nM Shh for 1 h. The interaction between β-arrestin1-Myc and Smo-GFP increased upon Shh treatment. D, Smo-GFP was coexpressed with β-arrestin1-Myc and Fyn-GST in Cos7 cells. Smo-GFP was immunoprecipitated. Fyn-GST only coimmunoprecipitated with Smo-GFP when β-arrestin1-Myc was present. E, Commissural neuron lysate was immunoprecipitated with an IgG isotype control or anti-Src antibody. Smo coimmunoprecipitated with Src, but not with the IgG control.

We then tested the interaction between Smo and β-arrestins by doing coimmunoprecipitation assays in Cos7 cells in which we expressed β-arrestin1/2-Myc with Smo-GFP or Smo-GFP alone. Both β-arrestins coimmunoprecipitated with Smo-GFP, indicating that β-arrestins can interact with Smo, and this was specific since there was no Myc signal in the coimmunoprecipitate when Smo-GFP, β-arrestin1-Myc, or β-arrestin2-Myc were expressed alone (Fig. 2B). Our result confirms the previously described interaction between Smo and β-arrestins (W. Chen et al., 2004; Kovacs et al., 2008). We also found that Shh stimulation increased the amount of β-arrestin1-Myc which coimmunoprecipitated with Smo-GFP (Fig. 2C). Previous studies reported a similar increased interaction between Smo and β-arrestins when cells were treated with Shh (W. Chen et al., 2004; Kovacs et al., 2008; Y. Chen et al., 2011).

Given that β-arrestins can interact with SFKs (Fig. 2A) and that β-arrestins can interact with Smo (Fig. 2B), we decided to determine whether β-arrestins can form a complex with both SFKs and Smo. We expressed Smo-GFP, Fyn-GST, or β-arrestin1-Myc either alone or in combination in Cos7 cells. When only Fyn-GST was coexpressed with Smo-GFP, immunoprecipitation of Smo-GFP did not result in the coimmunoprecipitation of Fyn-GST (Fig. 2D). When only β-arrestin1-Myc was coexpressed with Smo-GFP, β-arrestin1-Myc coimmunoprecipitated with Smo-GFP, similar to what we showed previously (Fig. 2B). Only when all three proteins were coexpressed together that immunoprecipitation of Smo-GFP resulted in the coimmunoprecipitation of Fyn-GST, together with β-arrestin1-Myc. Thus, Smo, β-arrestin1, and Fyn can form a complex, with β-arrestin1 mediating the interaction between Smo and Fyn.

To test whether endogenous Smo and the SFK Src can also form a complex in commissural neurons, we immunoprecipitated Src from commissural neuron lysates with an anti-Src antibody. We found that Smo coimmunoprecipitated with Src and was absent in the immunoprecipitation with the isotype control (Fig. 2E). This demonstrates that endogenous Smo and Src interact in commissural neurons.

β-arrestins are required for Shh-mediated activation of Src

Shh-mediated axon guidance requires SFK activation downstream of Smo activation (Yam et al., 2009). Since β-arrestins recruit SFKs to Smo, we hypothesized that β-arrestins are required for the activation of SFKs. To test this, we used siRNAs to knock down endogenous β-arrestins in commissural neurons. We designed two siRNAs directed against β-arrestin1 and β-arrestin2, respectively, and used them in combination to deplete both β-arrestins. We targeted both β-arrestins to ensure no compensatory mechanism by one β-arrestin for the other (DeWire et al., 2007). We first assessed the knockdown efficiency by Western blotting of commissural neuron lysate using an antibody detecting both β-arrestin1 and 2 (pan-arrestin; Fig. 3A). Commissural neurons electroporated with siRNAs against β-arrestin1/2 show a decreased β-arrestins protein level of 70% compared with neurons electroporated with nontargeting siRNA [scrambled (scr) siRNA; Fig. 3B]. We then determined the effect of β-arrestin knockdown on SFK activation by stimulating commissural neurons knocked down for β-arrestins with Shh and by detecting the level of phosphorylated Src at tyrosine residue 418 (pSrc Y418), the active form of Src (Roskoski, 2004). In commissural neurons electroporated with scr siRNA, the pSrc Y418 level increases upon Shh stimulation (Fig. 3C,D), as we had previously shown (Yam et al., 2009). However, when β-arrestins were knocked down, the Shh-induced increase in pSrc Y418 was abolished (Fig. 3E,F). These results indicate that β-arrestins are required for the activation of the SFK Src upon Shh stimulation.

Figure 3.
Figure 3.

β-arrestins are required for Shh-mediated activation of Src. A, Representative Western blot of β-arrestin1/2 in commissural neurons electroporated with a nontargeting siRNA (scrambled, scr) or siRNA targeting β-arrestin1/2. B, The relative amount (mean ± SEM) of β-arrestin1/2 normalized to the amount of GAPDH in the cell lysate. β-arrestin1/2 siRNA efficiently knocked down expression of β-arrestin1/2. n = 4 experiments, paired t test, **p value 0.0023. C, Representative Western blot of pSrc Y418 in commissural neurons electroporated with scr siRNA and stimulated with 0.1 µg/ml BSA or Shh for 30 min. D, The relative pSrc Y418 band intensity, normalized to GAPDH (mean ± SEM). In control (scr siRNA) electroporated commissural neurons, Shh stimulation increases pSrc Y418 levels. n = 3 experiments, paired t test, **p value 0.0022. E, Representative Western blot of pSrc Y418 in commissural neurons electroporated with β-arrestin1/2 siRNA and stimulated with BSA or Shh. F, The relative pSrc Y418 band intensity, normalized to GAPDH (mean ± SEM). When β-arrestin1/2 are knocked down, Shh stimulation fails to increase pSrc Y418 levels. n = 3 experiments, paired t test, ns, not significant.

β-arrestins are required for Shh-mediated commissural axon attraction

We tested the importance of β-arrestins for commissural axon attraction using Dunn chamber turning assays, in which we can expose commissural neurons to a gradient of axon guidance molecules and measure their attraction up the gradient, i.e., toward higher concentrations of the guidance molecules (Fig. 4A). We exposed commissural neurons electroporated with scr siRNA or with β-arrestin1/2 siRNA to a gradient of Shh or a control gradient and measured the turned angle of axons to see if they were attracted toward the gradient. Axons of control neurons (scr) exposed to a control gradient were not attracted up the gradient and did not change their direction of growth, as demonstrated by a mean turned angle close to 0° (Fig. 4B,C). When control neurons were exposed to a Shh gradient, their axons were attracted toward the Shh gradient, represented by a positive turned angle. However, when commissural neurons were knocked down for β-arrestins, axons were no longer attracted toward the Shh gradient (Fig. 4B,C). We observed no difference in the growth of axons from commissural neurons in the three different conditions, indicating that knockdown of β-arrestins did not affect axon extension and that the lack of axon turning toward Shh in commissural neurons knocked down for β-arrestins was not due to a lack of axon growth (Fig. 4D). Our results show that β-arrestins are required for commissural axon attraction toward Shh.

Figure 4.
Figure 4.

β-arrestins are required for Shh-mediated commissural axon attraction. A, Top and side view of a Dunn chamber, from Yam et al. (2009). Diffusion of a chemoattractant from the outer well to the inner well forms a gradient over the bridge. Axons that are exposed to a Shh gradient formed over the bridge are imaged and analyzed for their ability to be attracted by Shh. B, Commissural axons electroporated with scr siRNA and exposed to a control gradient do not turn toward the control gradient but are attracted toward a Shh gradient. In contrast, commissural neurons knocked down for β-arrestin1/2 are not attracted up a Shh gradient. Scale bar, 20 µm. C, Quantification of the mean angle turned (mean ± SEM), positive angles representing attraction, and negative angles indicating repulsion. D, Net axon growth (mean ± SEM) is not affected by β-arrestin1/2 knockdown. C, D, n = 8 independent experiments, 8–72 neurons analyzed per condition per experiment, one-way ANOVA, Tukey's multiple-comparisons test *p value 0.0313, **p value 0.0099, ns, not significant.

The Smo internalization and SFK binding functions of β-arrestins are required for Shh-mediated commissural axon attraction

Since we determined that β-arrestins are required for Shh-mediated attraction of commissural axons, we next explored which functions of β-arrestins are important for axon guidance. We took advantage of already characterized dominant-negative (DN) β-arrestin1 constructs. V53D β-arrestin1 is still able to interact with SFKs but is impaired in its capacity to mediate the internalization of GPCRs (Ferguson et al., 1996; Luttrell et al., 1999; Vögler et al., 1999). Since β-arrestins also mediate the internalization of Smo (W. Chen et al., 2004), we reasoned that V53D β-arrestin1 may also block the internalization of Smo, as it does with other GPCRs such as the β2AR and muscarinic acetylcholine receptors (Ferguson et al., 1996; Vögler et al., 1999). To test this, we coexpressed Smo-GFP with either an empty vector, WT β-arrestin1-Myc, or V53D β-arrestin1-Myc in NIH3T3 cells and looked at the localization of Smo-GFP by immunofluorescence microscopy. We observed two different patterns of Smo-GFP localization: (1) a diffuse expression of Smo-GFP throughout the cytoplasm and absence of Smo-GFP puncta or (2) the presence of bright Smo-GFP puncta in the cytoplasm (Fig. 5A). We found that expression of WT β-arrestin1-Myc significantly increased the percentage of cells with Smo-GFP puncta, compared with the empty vector control (Fig. 5B). This is consistent with the known role of β-arrestin for the internalization of Smo, where induction of Smo internalization to clathrin-coated pits is associated with an increase in large intracellular Smo puncta (W. Chen et al., 2004). Therefore, the increase in Smo-GFP puncta upon WT β-arrestin1-Myc expression most likely reflects increased internalized Smo-GFP (Fig. 5C). When we expressed V53D β-arrestin1 in NIH3T3 cells, we found that the proportion of cells presenting a punctate Smo-GFP localization was not significantly different from the cells expressing the empty vector control (Fig. 5B). Given that V53D β-arrestin1 blocks the internalization of other GPCRs and that it lacks the ability to increase Smo-GFP puncta formation in the cytoplasm, it suggests that V53D β-arrestin1 impairs Smo-GFP trafficking, most likely at the level of endocytosis (Fig. 5C; Luttrell et al., 1999). The other dominant-negative construct, P91G-P121E β-arrestin1, has an altered SH3 domain-binding region that is essential for its interaction with SFKs. Thus it has a reduced interaction with Src but can still interact with GPCRs and mediate GPCR internalization (Luttrell et al., 1999). We tested if in addition to a loss of interaction with Src, P91G-P121E β-arrestin1 is also unable to interact with Fyn. We coexpressed Fyn-GST with either WT, V53D, or P91G-P121E β-arrestin1-Myc in Cos7 cells. WT β-arrestin1 and V53D β-arrestin1, but not P91G-P121E β-arrestin1, were pulled down by Fyn-GST (Fig. 5D). This demonstrates that P91G-P121E β-arrestin1, in addition to not interacting with Src, also does not interact with Fyn (Fig. 5E), whereas the V53D β-arrestin1 which affected Smo internalization (Fig. 5A–C) is still able to interact with Fyn.

Figure 5.
Figure 5.

Dominant-negative β-arrestin1 inhibits Shh-mediated commissural axon attraction. A, NIH3T3 cells were cotransfected with Smo-GFP and either an empty vector control, WT β-arrestin1-Myc, or V53D β-arrestin1-Myc. After 48 h, cells were fixed and Smo-GFP was detected with an anti-GFP antibody. Cells were classified according to the presence or absence of Smo-GFP puncta in the cytosol. Representative images of NIH3T3 cells expressing Smo-GFP illustrating the categories “puncta absent” and “puncta present” used for classification. Red arrows indicate Smo-GFP puncta. Scale bar: 20 µm (left), 10 µm in zoomed image (right). B, Percentage of cells (mean ± SEM) with Smo-GFP puncta. Empty vector condition n = 2 experiments, WT β-arrestin1 and V53D β-arrestin1, n = 3 experiments, 13–30 cells analyzed per condition per experiment, one-way ANOVA, *p value 0.0195, **p value 0.0032, ns, not significant. C, Wild-type β-arrestin1 (WT β-arr1) promotes Smo internalization whereas dominant-negative (DN) V53D β-arrestin1 (V53D β-arr1) does not. D, Fyn-GST was coexpressed with WT β-arrestin1-Myc or DN β-arrestin1-Myc in Cos7 cells. Fyn-GST was PD from cell lysate using glutathione beads and β-arrestin1-Myc was detected with an anti-Myc antibody. Only P91G-P121E β-arrestin1-Myc was not pulled-down with Fyn-GST, indicating that it is not able to interact with Fyn-GST. E, WT β-arrestin1 interacts with Smo and SFKs and DN P91G-P121E β-arrestin1 (P91G-P121E β-arr1) does not interact with SFKs but can still interact with Smo. F, Commissural axons were electroporated with GFP control, WT β-arrestin1-Myc, or DN WT β-arrestin1-Myc constructs and exposed to a Shh gradient in a Dunn chamber. Quantification of the mean angle turned (mean ± SEM), positive angles representing attraction, and negative angles indicating repulsion. Commissural axons expressing GFP or WT β-arrestin1 are attracted toward Shh. Expression of either DN β-arrestin1 blocks commissural axon attraction toward Shh. G, Expression of DN β-arrestin1 does not affect the axon growth of commissural neurons (mean ± SEM). F, G, n = 6 independent experiments, 18–48 neurons per condition per experiment, one-way ANOVA, Tukey's multiple-comparisons test. F, G, *p value <0.05, **p value 0.0012, ns, not significant.

Figure 5-1

Expression of DN β-arrestin1 in commissural neurons. Commissural neurons expressing WT β-arr1-Myc, P91G-P121E β-arr1-Myc or V53D β-arr1-Myc, as detected by immunofluorescence staining using an anti-Myc antibody. Scale = 100  µm (left) or 20  µm (right). Download Figure 5-1, TIF file.

Figure 5-2

Expression of DN β-arrestin1 does not affect the viability and health of commissural neurons. Commissural neurons were electroporated with GFP control, WT β-arrestin1-Myc or DN WT β-arrestin1-Myc constructs. After two days of culture, the neurons were fixed and immunostained for cleaved caspase-3 to detect apoptotic cells and GFP or Myc to detect transfected cells. (A) Percentage (mean ± SEM) of commissural neurons expressing GFP or β-arrestin1-Myc constructs that are also positive for cleaved caspase-3. n = 1 experiment (GFP, 185 cells counted), n = 2 experiments (WT, V53D and P91G-P121E β-arrestin1, 155-265 cells counted p­er experiment). (B) Axon length (mean ± SEM) of commissural neurons expressing GFP or β-arrestin1-Myc constructs. 106-116 axons per condition. Download Figure 5-2, TIF file.

We next expressed these DN β-arrestin1 constructs in commissural neurons and assessed the impact of their expression on the ability of neurons to respond to a Shh gradient in Dunn chambers. All three β-arrestin1-Myc constructs (WT β-arrestin1, V53D β-arrestin1, and P91G-P121E β-arrestin1) could be expressed in commissural neurons and were present throughout the neuron, including the growth cone (Extended Data Fig. 5-1). Expression of the various β-arrestin1-Myc constructs did not affect the viability of the neurons as assessed by cleaved caspase-3 immunostaining (Extended Data Fig. 5-2A). Expression of the various β-arrestin1-Myc constructs also had no effect on total axon length (Extended Data Fig. 5-2B), suggesting that overall neuron health was unaffected. Axons from commissural neurons expressing GFP or WT β-arrestin1 were attracted toward a Shh gradient, indicated by the positive angle turned (Fig. 5F). In contrast, axons from neurons expressing either V53D β-arrestin1 or P91G-P121E β-arrestin1 were unable to turn toward the Shh gradient (Fig. 5F). This indicates that Shh-mediated axon attraction requires the Smo internalization function and the SFK-binding function of β-arrestins. Expression of the different constructs did not affect axon growth (Fig. 5G), consistent with the constructs not affecting total axon length (Extended Data Fig. 5-2B), indicating that DN β-arrestin1 did not cause deleterious effects on the neurons and that the lack of axon turning toward Shh when DN β-arrestin1 was expressed was not due to a lack of axon growth. Moreover, that expression of DN β-arrestin1 inhibited Shh-mediated commissural axon attraction, similar to the effect of siRNA knockdown of β-arrestins, indicates that the effect of the β-arrestins siRNA was not due to an off-target effect.

β-arrestins are required for commissural axon guidance in vivo

After demonstrating that β-arrestins are required for Shh-mediated attraction of commissural axons, we next sought to test whether β-arrestins are also required for axon guidance in vivo. We coelectroporated HH stage 11–12 chick embryo neural tubes with a control plasmid, WT β-arrestin1-Myc, V53D β-arrestin1-Myc, or P91G-P121E β-arrestin1-Myc together with a J2X-GFP plasmid. We did not observe changes in the morphology of the spinal cord under the different conditions. The J2X-GFP plasmid expresses GFP under the Math1 promoter, which is expressed in the dorsal population of commissural neurons (Helms and Johnson, 1998), allowing us to visualize electroporated commissural axons by GFP expression. We also visualized the expression of the Myc-tagged β-arrestin1 constructs in the spinal cord by immunostaining for Myc. In the control electroporated neural tubes, commissural axons follow a normal trajectory, extending from the dorsal to the ventral spinal cord without entering the motor column. Commissural axons from spinal cords expressing WT β-arrestin1 also follow a typical trajectory and no axons deviate from the main axon bundle (Fig. 6A). The expression of either V53D β-arrestin1 or P91G-P121E β-arrestin1 caused axons to deviate from the main axon bundle (Fig. 6A). We observed axons invading the lateral ventral spinal cord and motor column. We calculated the widening of the axon bundle by measuring the area occupied by axons expressing GFP over the total ventral area of the spinal cord. Our measurements indicate that commissural axons from spinal cords expressing DN β-arrestin1 are more dispersed compared with both control spinal cord and spinal cord expressing WT β-arrestin1 (Fig. 6B,C). This demonstrates that β-arrestins are required for correct commissural axon guidance in vivo. Notably, the defects we see when β-arrestins are disrupted phenocopies the defects seen in mutants of other proteins involved in Shh-mediated axon guidance, such as Boc, conditional deletion of Smo, and ZBP1 (Charron et al., 2003; Okada et al., 2006; Lepelletier et al., 2017). This supports β-arrestins having a function in the noncanonical Shh signaling pathway. Together with our in vitro data demonstrating that β-arrestins are required for Shh-mediated axon attraction, we conclude that β-arrestins are required for Shh-mediated axon guidance in vitro and in vivo.

Figure 6.
Figure 6.

Dominant-negative β-arrestin1 impair commissural axon guidance in vivo. A, Chick embryo neural tubes were coelectroporated with J2X-EGFP and WT β-arrestin1-Myc or DN β-arrestin1-Myc at HH stage 11–12 and fixed and immunostained at HH stage 28–29. Commissural neurons were labeled with GFP and β-arrestin1 expression was detected by an anti-Myc antibody. Arrowheads delimit the main bundle of commissural axons. Pre-crossing commissural axon tracts appear normal in control and in WT β-arrestin1 electroporated neural tubes. Expression of DN β-arrestin1 results in a wider commissural axon tract and commissural axons deviating from the main bundle, invading the ventral spinal cord and motor column. Scale bar, 200 µm. B, The ventral area occupied by GFP-positive axons (mean ± SEM) over the total ventral area of the spinal cord. n ≥ 4 embryos per condition, one-way ANOVA, Tukey's multiple-comparisons test. *p value 0.0164, **p values <0.01, ***p value 0.0003, ns, not significant. C, Schematic depicting commissural axons (in green) in embryos expressing control or WT β-arrestin1 (left) and in embryos expressing DN β-arrestin1 (right). Commissural axons follow a stereotypical trajectory in embryos expressing control and WT β-arrestin1, whereas commissural axons invade the motor column and have a wider trajectory in the embryos expressing DN β-arrestin1.

Discussion

By virtue of their ability to bind to a variety of GPCRs to induce receptor desensitization, receptor internalization or activation of intracellular signaling pathways, β-arrestins have multiple roles in the nervous system. For example, β-arrestin-mediated desensitization of olfactory receptors enables discrimination between intraneuronal olfactory stimuli (Merritt et al., 2022), β-arrestin-dependent Frizzled endocytosis is important for Caenorhabditis elegans PLM mechanosensory neuron branching (C-H. Chen et al., 2017), β-arrestin-dependent signaling is activated by D3 dopamine receptors to regulate axon initial segment excitability (S. Yang et al., 2016), and β-arrestin2 is required for NMDA-induced remodeling of dendritic spines via translocation of active cofilin (Pontrello et al., 2012).

Our work identifies β-arrestins as scaffolding proteins required for axon guidance. We found that SFKs can be recruited to Smo in a β-arrestin-dependent manner, with β-arrestins forming a tripartite complex with Smo and SFKs. Endogenous SFK and Smo proteins also interact in commissural neurons. We also found that Shh stimulation recruits β-arrestins to Smo. We show that β-arrestins are required for SFK activation in response to Shh and for axon attraction by Shh (Fig. 7). The expression of DN V53D β-arrestin1 and P91G-P121E β-arrestin1 in commissural neurons also prevented Shh-mediated attraction of axons in vitro and caused a widening of commissural axon tracts in vivo, consistent with the Smo internalization and SFK binding functions of β-arrestin being required for Shh-mediated axon guidance. Together our results support a model where the binding of Shh to Boc and Ptch1 relieves inhibition of Smo. Activated Smo recruits β-arrestin1/2 which interacts with and activates SFKs. Our model attributes a scaffolding role for β-arrestins in Shh-mediated axon guidance, where β-arrestins recruit SFKs to Smo for the activation of SFKs required for axon attraction (Fig. 7).

Figure 7.
Figure 7.

β-arrestins are scaffolding proteins required for Shh-mediated axon turning. Binding of Shh to Boc and Ptch1 leads to the activation of Smo. β-arrestins are recruited to activated Smo and bind to SFKs, forming a tripartite complex. This leads to the phosphorylation and activation of SFKs and axon turning.

Smo activation recruits β-arrestins

The interaction between Smo and β-arrestin is increased by Shh stimulation (Kovacs et al., 2008) and inhibited by a Smo antagonist (Kovacs et al., 2008), indicating that Smo activation recruits β-arrestins and that Smo inactivation prevents the binding of β-arrestins. In mammalian cells, GRK2 phosphorylates Smo and promotes the recruitment of β-arrestin2 to Smo. Smo phosphorylation by GRK2 is also increased by a Smo agonist and decreased by a Smo antagonist (W. Chen et al., 2004). Shh induces Smo phosphorylation at its carboxyl-tail (C-tail) by acting through GRK2 and Casein kinase 1α (CK1α; Y. Chen et al., 2011). Therefore we propose that upon Shh stimulation, Smo activation leads to phosphorylation of the Smo C-tail, favoring the recruitment of β-arrestins to Smo, a typical mechanism of recruitment of β-arrestins to GPCRs (Evron et al., 2012; Arensdorf et al., 2016; Gurevich and Gurevich, 2019).

β-arrestin-mediated Smo internalization is important for Shh-mediated axon guidance

In response to Shh stimulation, Smo is internalized from the cell membrane and segregates into endosomes (Incardona et al., 2002). In mammalian cells, β-arrestin internalizes Smo, resulting in large intracellular Smo puncta in the cytoplasm (W. Chen et al., 2004). In Drosophila, expression of Kurtz, the Drosophila β-arrestin homolog, reduces Smo levels at cell membranes (Cheng et al., 2010; Molnar et al., 2011). The V53D β-arrestin1 mutant acts as a DN for GPCR internalization, impairing the endocytosis of β2AR (Ferguson et al., 1996) and muscarinic acetylcholine receptors (Vögler et al., 1999). We found that unlike WT β-arrestin, which increases cytoplasmic Smo puncta formation, DN V53D β-arrestin1 is unable to stimulate Smo puncta formation. Likewise, the V94D mutant in Kurtz, orthologous to V53D β-arrestin1, is also unable to affect Drosophila Smo membrane localization (Molnar et al., 2011). Altogether this is consistent with V53D β-arrestin1/V94D Kurtz being unable to support Smo internalization.

We found that the expression of DN V53D β-arrestin1 was sufficient to abolish the axon turning response toward Shh in vitro and disrupt commissural axon guidance in vivo. This suggests that Smo internalization via β-arrestins is required for Shh-mediated axon guidance.

The regulation and role of Smo internalization for Shh signaling is not well understood. In mammalian cells, the phosphorylation of Smo by GRK2 recruits β-arrestin which mediates clathrin-mediated endocytosis of Smo (W. Chen et al., 2004). Smo internalization may be necessary for Smo trafficking to other cellular locales for signaling, or it may be important for the recycling of Smo to the plasma membrane. In canonical Shh signaling, β-arrestins are required for the accumulation of Smo at the primary cilia in response to Shh (Kovacs et al., 2008). This raises the possibility that endocytosis of Smo and its subsequent trafficking may be required to generate vesicles that carry Smo to the cilia (Kovacs et al., 2008; Rohatgi and Scott, 2008). The internalization of Smo may even be important for certain forms of Smo activity, as suggested by the identification of an intracellular active form of Smo (Jiang et al., 2018).

The endocytosis of GPCRs mediated by β-arrestins is important for the desensitization of the receptors and their recycling back to the plasma membrane to resensitize the pathway (Ayers and Thérond, 2010). This process could explain why Smo internalization is necessary for Shh-mediated attraction and canonical Shh signaling. In the presence of Shh, the binding of β-arrestins to Smo would promote its internalization and would lead to the rapid recycling of Smo back to the plasma membrane. This fast recycling of Smo would allow it to continue transducing the Shh signal, hence maintaining the responsiveness to Shh for axon guidance. Our group reported that Shh stimulation increases Smo surface localization in commissural neurons (Ferent et al., 2019). This increase in Smo at the surface of the cell could be the consequence of the fast recycling of Smo at the membrane mediated by β-arrestins. Alternatively, Smo internalization may be important for its ability to form a complex with and activate SFKs in axon guidance, similarly to how GPCR internalization is required for formation of a signalosome and ERK1/2 activation (Luttrell et al., 1999).

β-arrestins mediate SFK activation for Shh-mediated axon attraction

β-arrestins (this study) and Smo (Yam et al., 2009) are both required for SFK activation downstream of Shh. Indeed, the Shh-induced activation of Src characterized by phosphorylation at tyrosine 418 is lost when we knocked down β-arrestins from commissural neurons. Moreover, P91G-P121E β-arrestin1, which is unable to interact with SFKs, is unable to mediate Shh-mediated attraction, suggesting that the interaction between β-arrestins and SFKs is important for Shh-mediated attraction, possibly because SFKs are not activated without this interaction. Recent work has identified a mechanism for the activation of Src downstream of GPCRs. This mechanism relies on the interaction of β-arrestin1 with the phosphorylated GPCR tail. The interaction between the GPCR-β-arrestin1 complex and Src promotes Src autophosphorylation of tyrosine 416 in chick (orthologous to tyrosine 418 in mammals), activating the kinase (Pakharukova et al., 2020). The same process may occur with Smo and β-arrestins, where β-arrestins bind to Smo and then recruit Src for its activation. SFKs are necessary for Shh-mediated commissural axon guidance (Yam et al., 2009), and SFKs phosphorylate many different proteins acting in cytoskeleton remodeling, important for axon guidance (Knöll and Drescher, 2004; Liu et al., 2007; Bashaw and Klein, 2010). Thus, the Shh-dependent activation of SFKs by β-arrestins is critical for cytoskeleton remodeling and axon guidance.

Altogether, our data suggest that β-arrestins act as scaffold proteins in Shh-mediated axon guidance. The scaffolding role of β-arrestins is linked to the activation of many cellular pathways (Luttrell et al., 1999; Peterson and Luttrell, 2017), and our work highlights the role of β-arrestins in Shh signaling. Notably, although β-arrestins are required for both canonical and noncanonical Shh signaling, its role in the recruitment and activation of SFKs is specific to noncanonical Shh signaling. Our findings implicate β-arrestins in axon guidance, filling the gap between Smo activation and SFK activation, thus giving a better comprehension of noncanonical Shh signaling.

Footnotes

  • Work in the laboratory of F.C. is funded by the Canadian Institutes of Health Research grant PJT180647 and FDN334023 and the Canada Foundation for Innovation grants CFI33768 and CFI39794. F.C. holds the Canada Research Chair in Developmental Neurobiology. We thank J.-F. Michaud, N. Balekoglu, S. Schlienger, and K. Zhang for technical assistance and S. Calabretta for critical feedback on the manuscript. We are grateful to A. Veillette for providing the anti-Src antibody, M. Bouvier for providing pcDNA3-β-arrestin1-WT-Myc, and R. Screaton for providing pDEST27 Fyn-GST.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Patricia T. Yam at patricia.yam{at}ircm.qc.ca or Frédéric Charron at frederic.charron{at}ircm.qc.ca.

References

  1. ↵
    1. Alexander SP,
    2. Christopoulos A,
    3. Davenport AP,
    4. Kelly E,
    5. Marrion NV,
    6. Peters JA,
    7. Faccenda E,
    8. Harding SD,
    9. Pawson AJ,
    10. Sharman JL
    (2017) The concise guide to pharmacology 2017/18: G protein-coupled receptors. Br J Pharmacol 174:S17–S129. https://doi.org/10.1111/bph.13878 pmid:29055040
    OpenUrlCrossRefPubMed
  2. ↵
    1. Allen BL,
    2. Song JY,
    3. Izzi L,
    4. Althaus IW,
    5. Kang J-S,
    6. Charron F,
    7. Krauss RS,
    8. McMahon AP
    (2011) Overlapping roles and collective requirement for the coreceptors GAS1, CDO, and BOC in SHH pathway function. Dev Cell 20:775–787. https://doi.org/10.1016/j.devcel.2011.04.018 pmid:21664576
    OpenUrlCrossRefPubMed
  3. ↵
    1. Arensdorf AM,
    2. Marada S,
    3. Ogden SK
    (2016) Smoothened regulation: a tale of two signals. Trends Pharmacol Sci 37:62–72. https://doi.org/10.1016/j.tips.2015.09.001 pmid:26432668
    OpenUrlCrossRefPubMed
  4. ↵
    1. Ayers KL,
    2. Thérond PP
    (2010) Evaluating smoothened as a G-protein-coupled receptor for hedgehog signalling. Trends Cell Biol 20:287–298. https://doi.org/10.1016/j.tcb.2010.02.002
    OpenUrlCrossRefPubMed
  5. ↵
    1. Barlic J,
    2. Andrews JD,
    3. Kelvin AA,
    4. Bosinger SE,
    5. DeVries ME,
    6. Xu L,
    7. Dobransky T,
    8. Feldman RD,
    9. Ferguson SS,
    10. Kelvin DJ
    (2000) Regulation of tyrosine kinase activation and granule release through β-arrestin by CXCR1. Nat Immunol 1:227–233. https://doi.org/10.1038/79767
    OpenUrlCrossRefPubMed
  6. ↵
    1. Bashaw GJ,
    2. Klein R
    (2010) Signaling from axon guidance receptors. Cold Spring Harb Perspect Biol 2:a001941. https://doi.org/10.1101/cshperspect.a001941 pmid:20452961
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Charron F,
    2. Stein E,
    3. Jeong J,
    4. McMahon AP,
    5. Tessier-Lavigne M
    (2003) The morphogen sonic hedgehog is an axonal chemoattractant that collaborates with netrin-1 in midline axon guidance. Cell 113:11–23. https://doi.org/10.1016/S0092-8674(03)00199-5
    OpenUrlCrossRefPubMed
  8. ↵
    1. Chen W,
    2. Burgess S,
    3. Hopkins N
    (2001) Analysis of the zebrafish smoothened mutant reveals conserved and divergent functions of hedgehog activity. Development 128:2385–2396. https://doi.org/10.1242/dev.128.12.2385
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Chen C-H,
    2. He C-W,
    3. Liao C-P,
    4. Pan C-L
    (2017) A Wnt-planar polarity pathway instructs neurite branching by restricting F-actin assembly through endosomal signaling. PLoS Genet 13:e1006720. https://doi.org/10.1371/journal.pgen.1006720 pmid:28384160
    OpenUrlCrossRefPubMed
  10. ↵
    1. Chen W,
    2. Ren X-R,
    3. Nelson CD,
    4. Barak LS,
    5. Chen JK,
    6. Beachy PA,
    7. de Sauvage F,
    8. Lefkowitz RJ
    (2004) Activity-dependent internalization of smoothened mediated by β-arrestin 2 and GRK2. Science 306:2257–2260. https://doi.org/10.1126/science.1104135
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Chen Y,
    2. Sasai N,
    3. Ma G,
    4. Yue T,
    5. Jia J,
    6. Briscoe J,
    7. Jiang J
    (2011) Sonic hedgehog dependent phosphorylation by CK1α and GRK2 is required for ciliary accumulation and activation of smoothened. PLoS Biol 9:e1001083. https://doi.org/10.1371/journal.pbio.1001083 pmid:21695114
    OpenUrlCrossRefPubMed
  12. ↵
    1. Cheng S,
    2. Maier D,
    3. Neubueser D,
    4. Hipfner DR
    (2010) Regulation of smoothened by Drosophila G-protein-coupled receptor kinases. Dev Biol 337:99–109. https://doi.org/10.1016/j.ydbio.2009.10.014 pmid:19850026
    OpenUrlCrossRefPubMed
  13. ↵
    1. DeWire SM,
    2. Ahn S,
    3. Lefkowitz RJ,
    4. Shenoy SK
    (2007) β-arrestins and cell signaling. Annu Rev Physiol 69:483–510. https://doi.org/10.1146/annurev.physiol.69.022405.154749
    OpenUrlCrossRefPubMed
  14. ↵
    1. Evron T,
    2. Daigle TL,
    3. Caron MG
    (2012) GRK2: multiple roles beyond G protein-coupled receptor desensitization. Trends Pharmacol Sci 33:154–164. https://doi.org/10.1016/j.tips.2011.12.003 pmid:22277298
    OpenUrlCrossRefPubMed
  15. ↵
    1. Ferent J,
    2. Giguere F,
    3. Jolicoeur C,
    4. Morin S,
    5. Michaud JF,
    6. Makihara S,
    7. Yam PT,
    8. Cayouette M,
    9. Charron F
    (2019) Boc acts via numb as a Shh-dependent endocytic platform for Ptch1 internalization and Shh-mediated axon guidance. Neuron 102:1157–1171.e1155. https://doi.org/10.1016/j.neuron.2019.04.003
    OpenUrlCrossRef
  16. ↵
    1. Ferguson SS
    (2001) Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signaling. Pharmacol Rev 53:1–24.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Ferguson SS,
    2. Downey WE,
    3. Colapietro A-M,
    4. Barak LS,
    5. Ménard L,
    6. Caron MG
    (1996) Role of β-arrestin in mediating agonist-promoted G protein-coupled receptor internalization. Science 271:363–366. https://doi.org/10.1126/science.271.5247.363
    OpenUrlAbstract
  18. ↵
    1. Gurevich VV,
    2. Gurevich EV
    (2019) GPCR signaling regulation: the role of GRKs and arrestins. Front Pharmacol 10:125. https://doi.org/10.3389/fphar.2019.00125 pmid:30837883
    OpenUrlCrossRefPubMed
  19. ↵
    1. Helms AW,
    2. Johnson JE
    (1998) Progenitors of dorsal commissural interneurons are defined by MATH1 expression. Development 125:919–928. https://doi.org/10.1242/dev.125.5.919
    OpenUrlAbstract
  20. ↵
    1. Incardona JP,
    2. Gruenberg J,
    3. Roelink H
    (2002) Sonic hedgehog induces the segregation of patched and smoothened in endosomes. Curr Biol 12:983–995. https://doi.org/10.1016/S0960-9822(02)00895-3
    OpenUrlCrossRefPubMed
  21. ↵
    1. Ingham PW,
    2. McMahon AP
    (2001) Hedgehog signaling in animal development: paradigms and principles. Genes Dev 15:3059–3087. https://doi.org/10.1101/gad.938601
    OpenUrlFREE Full Text
  22. ↵
    1. Izzi L,
    2. Levesque M,
    3. Morin S,
    4. Laniel D,
    5. Wilkes BC,
    6. Mille F,
    7. Krauss RS,
    8. McMahon AP,
    9. Allen BL,
    10. Charron F
    (2011) Boc and Gas1 each form distinct Shh receptor complexes with Ptch1 and are required for Shh-mediated cell proliferation. Dev Cell 20:788–801. https://doi.org/10.1016/j.devcel.2011.04.017 pmid:21664577
    OpenUrlCrossRefPubMed
  23. ↵
    1. Jiang K,
    2. Liu Y,
    3. Zhang J,
    4. Jia J
    (2018) An intracellular activation of smoothened that is independent of hedgehog stimulation in Drosophila. J Cell Sci 131:jcs211367. https://doi.org/10.1242/jcs.211367 pmid:29142103
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Kania A,
    2. Jessell TM
    (2003) Topographic motor projections in the limb imposed by LIM homeodomain protein regulation of ephrin-A:EphA interactions. Neuron 38:581–596. https://doi.org/10.1016/S0896-6273(03)00292-7
    OpenUrlCrossRefPubMed
  25. ↵
    1. Karlstrom RO,
    2. Tyurina OV,
    3. Kawakami A,
    4. Nishioka N,
    5. Talbot WS,
    6. Sasaki H,
    7. Schier AF
    (2003) Genetic analysis of zebrafish gli1 and gli2 reveals divergent requirements for gli genes in vertebrate development. Development 130:1549–1564. https://doi.org/10.1242/dev.00364
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Knöll B,
    2. Drescher U
    (2004) Src family kinases are involved in EphA receptor-mediated retinal axon guidance. J Neurosci 24:6248–6257. https://doi.org/10.1523/JNEUROSCI.0985-04.2004 pmid:15254079
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Kovacs JJ,
    2. Whalen EJ,
    3. Liu R,
    4. Xiao K,
    5. Kim J,
    6. Chen M,
    7. Wang J,
    8. Chen W,
    9. Lefkowitz RJ
    (2008) β-Arrestin–mediated localization of smoothened to the primary cilium. Science 320:1777–1781. https://doi.org/10.1126/science.1157983 pmid:18497258
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Langlois SD,
    2. Morin S,
    3. Yam PT,
    4. Charron F
    (2010) Dissection and culture of commissural neurons from embryonic spinal cord. J Vis Exp 39:e1773. https://doi.org/10.3791/1773 pmid:20505653
    OpenUrlPubMed
  29. ↵
    1. Lefkowitz RJ,
    2. Shenoy SK
    (2005) Transduction of receptor signals by β-arrestins. Science 308:512–517. https://doi.org/10.1126/science.1109237
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Lepelletier L,
    2. Langlois SD,
    3. Kent CB,
    4. Welshhans K,
    5. Morin S,
    6. Bassell GJ,
    7. Yam PT,
    8. Charron F
    (2017) Sonic hedgehog guides axons via zipcode binding protein 1-mediated local translation. J Neurosci 37:1685–1695. https://doi.org/10.1523/JNEUROSCI.3016-16.2016 pmid:28073938
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Liu G,
    2. Li W,
    3. Gao X,
    4. Li X,
    5. Jürgensen C,
    6. Park H-T,
    7. Shin N-Y,
    8. Yu J,
    9. He M-L,
    10. Hanks SK
    (2007) p130CAS is required for netrin signaling and commissural axon guidance. J Neurosci 27:957–968. https://doi.org/10.1523/JNEUROSCI.4616-06.2007 pmid:17251438
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Luria V,
    2. Krawchuk D,
    3. Jessell TM,
    4. Laufer E,
    5. Kania A
    (2008) Specification of motor axon trajectory by ephrin-B:EphB signaling: symmetrical control of axonal patterning in the developing limb. Neuron 60:1039–1053. https://doi.org/10.1016/j.neuron.2008.11.011
    OpenUrlCrossRefPubMed
  33. ↵
    1. Luttrell L, et al.
    (1999) β-Arrestin-dependent formation of β2 adrenergic receptor-Src protein kinase complexes. Science 283:655–661. https://doi.org/10.1126/science.283.5402.655
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Makihara S,
    2. Morin S,
    3. Ferent J,
    4. Cote JF,
    5. Yam PT,
    6. Charron F
    (2018) Polarized dock activity drives Shh-mediated axon guidance. Dev Cell 46:410–425.e417. https://doi.org/10.1016/j.devcel.2018.07.007
    OpenUrlCrossRef
  35. ↵
    1. Ménard L,
    2. Ferguson SS,
    3. Zhang J,
    4. Lin F-T,
    5. Lefkowitz RJ,
    6. Caron MG,
    7. Barak LS
    (1997) Synergistic regulation of β2-adrenergic receptor sequestration: intracellular complement of β-adrenergic receptor kinase and β-arrestin determine kinetics of internalization. Mol Pharmacol 51:800–808. https://doi.org/10.1124/mol.51.5.800
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Merritt DM,
    2. MacKay-Clackett I,
    3. Almeida SM,
    4. Tran C,
    5. Ansar S,
    6. van der Kooy D
    (2022) Arrestin-mediated desensitization enables intraneuronal olfactory discrimination in Caenorhabditis elegans. Proc Natl Acad Sci U S A 119:e2116957119. https://doi.org/10.1073/pnas.2116957119 pmid:35878038
    OpenUrlPubMed
  37. ↵
    1. Molnar C,
    2. Ruiz-Gómez A,
    3. Martin M,
    4. Rojo-Berciano S,
    5. Mayor F Jr.,
    6. de Celis JF
    (2011) Role of the Drosophila non-visual ss-arrestin kurtz in hedgehog signalling. PLoS Genet 7:e1001335. https://doi.org/10.1371/journal.pgen.1001335 pmid:21437272
    OpenUrlCrossRefPubMed
  38. ↵
    1. Okada A,
    2. Charron F,
    3. Morin S,
    4. Shin DS,
    5. Wong K,
    6. Fabre PJ,
    7. Tessier-Lavigne M,
    8. McConnell SK
    (2006) Boc is a receptor for sonic hedgehog in the guidance of commissural axons. Nature 444:369–373. https://doi.org/10.1038/nature05246
    OpenUrlCrossRefPubMed
  39. ↵
    1. Pakharukova N,
    2. Masoudi A,
    3. Pani B,
    4. Staus DP,
    5. Lefkowitz RJ
    (2020) Allosteric activation of proto-oncogene kinase Src by GPCR–beta-arrestin complexes. J Biol Chem 295:16773–16784. https://doi.org/10.1074/jbc.RA120.015400 pmid:32978252
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Peterson YK,
    2. Luttrell LM
    (2017) The diverse roles of arrestin scaffolds in G protein–coupled receptor signaling. Pharmacol Rev 69:256–297. https://doi.org/10.1124/pr.116.013367 pmid:28626043
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Pontrello CG,
    2. Sun M-Y,
    3. Lin A,
    4. Fiacco TA,
    5. DeFea KA,
    6. Ethell IM
    (2012) Cofilin under control of β-arrestin-2 in NMDA-dependent dendritic spine plasticity, long-term depression (LTD), and learning. Proc Natl Acad Sci U S A 109:E442–E451. https://doi.org/10.1073/pnas.1118803109 pmid:22308427
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Rohatgi R,
    2. Scott MP
    (2008) Arrestin' movement in cilia. Science 320:1726–1727. https://doi.org/10.1126/science.1160448
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Roskoski R Jr.
    (2004) Src protein–tyrosine kinase structure and regulation. Biochem Biophys Res Commun 324:1155–1164. https://doi.org/10.1016/j.bbrc.2004.09.171
    OpenUrlCrossRefPubMed
  44. ↵
    1. Schaeren-Wiemers N,
    2. Gerfin-Moser A
    (1993) A single protocol to detect transcripts of various types and expression levels in neural tissue and cultured cells: in situ hybridization using digoxigenin-labelled cRNA probes. Histochemistry 100:431–440. https://doi.org/10.1007/BF00267823
    OpenUrlCrossRefPubMed
  45. ↵
    1. Vögler O,
    2. Nolte B,
    3. Voss M,
    4. Schmidt M,
    5. Jakobs KH,
    6. van Koppen CJ
    (1999) Regulation of muscarinic acetylcholine receptor sequestration and function by β-arrestin. J Biol Chem 274:12333–12338. https://doi.org/10.1074/jbc.274.18.12333
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Wilbanks AM,
    2. Fralish GB,
    3. Kirby ML,
    4. Barak LS,
    5. Li Y-X,
    6. Caron MG
    (2004) β-Arrestin 2 regulates zebrafish development through the hedgehog signaling pathway. Science 306:2264–2267. https://doi.org/10.1126/science.1104193
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Yam PT,
    2. Charron F
    (2020) Nonconventional axon guidance cues: Hedgehog, TGF-β/BMP, and Wnts in axon guidance. In: Cellular migration and formation of axons and dendrites (Rubenstein J, Rakic P, Chen B, Kwan KY, eds), pp 175–199. London: Elsevier.
  48. ↵
    1. Yam PT,
    2. Langlois SD,
    3. Morin S,
    4. Charron F
    (2009) Sonic hedgehog guides axons through a noncanonical, Src-family-kinase-dependent signaling pathway. Neuron 62:349–362. https://doi.org/10.1016/j.neuron.2009.03.022
    OpenUrlCrossRefPubMed
  49. ↵
    1. Yang S,
    2. Ben-Shalom R,
    3. Ahn M,
    4. Liptak AT,
    5. van Rijn RM,
    6. Whistler JL,
    7. Bender KJ
    (2016) β-Arrestin-dependent dopaminergic regulation of calcium channel activity in the axon initial segment. Cell Rep 16:1518–1526. https://doi.org/10.1016/j.celrep.2016.06.098 pmid:27452469
    OpenUrlCrossRefPubMed
  50. ↵
    1. Yang F,
    2. Xiao P,
    3. Qu C-X,
    4. Liu Q,
    5. Wang L-Y,
    6. Liu Z-X,
    7. He Q-T,
    8. Liu C,
    9. Xu J-Y,
    10. Li R-R
    (2018) Allosteric mechanisms underlie GPCR signaling to SH3-domain proteins through arrestin. Nat Chem Biol 14:876–886. https://doi.org/10.1038/s41589-018-0115-3
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
β-arrestins Are Scaffolding Proteins Required for Shh-Mediated Axon Guidance
Rachelle Sauvé, Steves Morin, Patricia T. Yam, Frédéric Charron
Journal of Neuroscience 24 July 2024, 44 (30) e0261242024; DOI: 10.1523/JNEUROSCI.0261-24.2024
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.