Arl13b regulates ciliogenesis and the dynamic localization of Shh signaling proteins

Abstract

Arl13b, a ciliary protein within the ADP-ribosylation factor family and Ras superfamily of GTPases, is required for ciliary structure but has poorly defined ciliary functions. In this paper, we further characterize the role of Arl13b in cilia by examining mutant cilia in vitro and determining the localization and dynamics of Arl13b within the cilium. Previously, we showed that mice lacking Arl13b have abnormal Sonic hedgehog (Shh) signaling; in this study, we show the dynamics of Shh signaling component localization to the cilium are disrupted in the absence of Arl13b. Significantly, we found Smoothened (Smo) is enriched in Arl13b-null cilia regardless of Shh pathway stimulation, indicating Arl13b regulates the ciliary entry of Smo. Furthermore, our analysis defines a role for Arl13b in regulating the distribution of Smo within the cilium. These results suggest that abnormal Shh signaling in Arl13b mutant embryos may result from defects in protein localization and distribution within the cilium.

FIGURE 1:

Tubulin modification defects in Arl13b^hnn mutant MEFs. (A and B) Immunofluorescence for acetylated α-tubulin (ac-tubulin; green) and glutamylated tubulin (red) in wild-type (A) and Arl13b^hnn (B) MEFs shows reduced staining for both antibodies in Arl13b^hnn MEFs. (C) Quantification of the average fluorescence intensity of acetylated α-tubulin and glutamylated tubulin fluorescence relative to background staining. Error bars are ± SEM. *p < 0.0001 using Student's t test.

FIGURE 2:

Arl13b regulates ciliary length. (A) Quantification of the percent of ciliated cells in wild-type and Arl13b^hnn MEFs. (B) Schematic of the Arl13b constructs that were transfected into MEFs. (C and D) Immunofluorescence for acetylated α-tubulin (ac-tubulin) in wild-type and Arl13b^hnn MEFs shows shortened cilia in Arl13b^hnn MEFs. (E–L) Immunofluorescence for GFP and acetylated α-tubulin in wild-type and Arl13b^hnn MEFs expressing GFP-tagged Arl13b constructs. (M) Quantification of ciliary length in MEFs with and without transfection. The constructs are separated by those that localize to cilia and those that do not. Error bars are ± SD. For untransfected wild-type vs. Arl13b^hnn MEFs, p < 0.0001. For wild-type vs. wild-type–overexpressing Arl13b-GFP and untagged Arl13b in the cilium, p < 0.0001; for Arl13b^hnn vs. Arl13b^hnn overexpressing Arl13b-GFP and untagged Arl13b, p < 0.0001.

FIGURE 3:

Ciliary Arl13b is TritonX-100 soluble. (A and C) Immunofluorescence in IMCD3 cells shows that Arl13b colocalizes with the ciliary marker acetylated α-tubulin (A) and does not colocalize with the basal body marker γ-tubulin (C). (B and D) Triton X-100 treatment results in the loss of a majority of Arl13b staining in the cilium, although axoneme staining with acetylated α-tubulin remains (B). (E and F) The known ciliary membrane protein SSTR3-GFP shows a similar loss of ciliary staining with TritonX-100 treatment (F).

FIGURE 4:

Arl13b-GFP dynamics reflect those of a ciliary membrane protein. The figure shows overexposed images for viewing purposes, but all intensities were measured without overexposure of pixels. (A and B) Photobleaching of a region of Arl13b-GFP (A) in the center of the cilium shows recovery dynamics similar to SSTR3-GFP (B). (C) Quantification of the recovery dynamics for SSTR3-GFP and Arl13b-GFP shows no significant difference in the recovery curve (prior to 5.9 s, p > 0.3 using Student's t test). The relative fluorescence intensity was quantified as the intensity in the bleached region relative to the whole cilium. (D to G) Photobleaching Arl13b-GFP (D) and SSTR3-GFP (E) in the whole cilium results in little fluorescence recovery (G). IFT88-YFP has a faster recovery rate in the cilium (F and G); p < 0.02 for both Arl13b-GFP and SSTR3-GFP, compared with IFT88-YFP at 29.9 s. (G) Quantification of fluorescence intensity is determined as the intensity of the bleached region (cilium) relative to an unbleached region in the field. All experiments were corrected for background. Error bars are ± SEM (C and G).

FIGURE 5:

IFT88 recovery is intact in Arl13b mutant cells. (A) Western blot showing knockdown of Arl13b in IMCD3 cells. (B) Quantification of the percentage of cells showing cilia stained with acetylated α-tubulin and Arl13b. (C) Quantification of ciliary length in knockdown cells compared with no knockdown. (D and E) Immunofluorescence for (D) Arl13b and (E) acetylated α-tubulin (ac-tubulin) costained with RFP in cells expressing the 504 knockdown construct. (F) Recovery of IFT88-EYFP in IMCD3 cells with and without knockdown of Arl13b. Error bars are ± SEM (C and F) and ± SD (B).

FIGURE 6:

Shh signaling component localization is disrupted in Arl13b^hnn MEFs. (A) Quantification of Gli-luciferase activity relative to Renilla luciferase. (B–M) Immunofluorescence for acetylated α-tubulin (green) and Gli2, Gli3, Sufu, Smo, and Ptch1 (red). In wild-type MEFs (B, D, F, H, J, and L) Gli2, Gli3, and Sufu are enriched after Shh-conditioned media treatment, while Ptch1 levels are reduced. Arl13b^hnn MEFs (C, E, G, I, K, and M) do not show significant enrichment of Gli2, Gli3, Sufu, or Smo, and Ptch1 staining is not significantly reduced after Shh-conditioned media treatment. Insets in (J–M) are showing Ptch1 and Smo staining alone in grayscale. (N) Quantification of average fluorescence intensity in the tip of the cilium (Gli2, Gli3, and Sufu) or the entire cilium (Ptch1 and Smo) relative to cell body staining. Data from all experiments are shown. p < 10⁻⁷ for wild-type MEFs with conditioned media vs. untreated wild-type MEFs using the Gli and Smo antibodies; p < 0.05 using the Ptch1 and Sufu antibodies; p < 0.04 for Smo staining in Arl13b^hnn MEFS with conditioned media vs. untreated Arl13b^hnn MEFS; p < 0.04 for Ptch1 intensities in untreated wild-type MEFs vs. untreated Arl13b^hnn MEFs.