Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2012 Feb;22(2):333-45.
doi: 10.1038/cr.2011.134. Epub 2011 Aug 16.

A SNX10/V-ATPase Pathway Regulates Ciliogenesis in Vitro and in Vivo

Affiliations
Free PMC article

A SNX10/V-ATPase Pathway Regulates Ciliogenesis in Vitro and in Vivo

Yanqun Chen et al. Cell Res. .
Free PMC article

Abstract

Sorting nexins (SNXs) are phosphoinositide-binding proteins implicated in the sorting of various membrane proteins in vitro, but the in vivo functions of them remain largely unknown. We reported previously that SNX10 is a unique member of the SNX family genes in that it has vacuolation activity in cells. We investigate the biological function of SNX10 by loss-of-function assay in this study and demonstrate that SNX10 is required for the formation of primary cilia in cultured cells. In zebrafish, SNX10 is involved in ciliogenesis in the Kupffer's vesicle and essential for left-right patterning of visceral organs. Mechanistically, SNX10 interacts with V-ATPase complex and targets it to the centrosome where ciliogenesis is initiated. Like SNX10, V-ATPase regulates ciliogenesis in vitro and in vivo and does so synergistically with SNX10. We further discover that SNX10 and V-ATPase regulate the ciliary trafficking of Rab8a, which is a critical regulator of ciliary membrane extension. These results identify an SNX10/V-ATPase-regulated vesicular trafficking pathway that is crucial for ciliogenesis, and reveal that SNX10/V-ATPase, through the regulation of cilia formation in various organs, play an essential role during early embryonic development.

Figures

Figure 1
Figure 1
SNX10 regulates ciliogenesis and left-right patterning in zebrafish. (A) Overexpression of GFP-tagged zebrafish SNX10a induces the accumulation of large vacuoles (arrow) in Hela cells. SNX10b or the GFP does not have this activity. (B) The embryonic expression pattern of SNX10a at 10-somite stage as detected by in situ hybridization. It is not spatially restricted. Arrow points to the Kupffer's vesicle. (C) RT-PCR analysis for the effect of splice-blocking morpholino of SNX10a (10a-SP). DNA-sequencing analysis reveals that the exon 3 of SNX10a is deleted in the majority of the PCR products from the 10a-SP-treated embryos (the lower band). The protein product of this abnormally spliced mRNA is expected to be non-functional. (D) The general morphology of SNX10a morphants at day 2. SNX10a morphants show mild cardiac edema and are slightly short. (E) Heart looping visualized by in situ hybridization with the cardiac-specific marker cmlc2 at day 2 (ventral view). In wild-type embryos, R loop is the predominant one. When left-right patterning is disrupted, heart looping can be reversed (L loop) or remain in the midline (no loop). (F) Expression pattern of southpaw (spaw) in embryos of 16-18 somite stages (dorsal view). (G) Statistical analysis of heart looping in day 2 embryos. Heart looping is randomized in the SNX10a morphants while the standard control morpholino or morpholinos to SNX10b does not affect heart looping. (H) Ciliogenesis in KV of 10-somite stage embryos. Cilia are visualized by immunostaining with the anti-acetylated α-tubulin antibody (green). Blue indicates the DAPI staining of nuclei. In SNX10a morphants, the average number of cilia per KV is reduced from 55 to 32.
Figure 2
Figure 2
SNX10 regulates ciliogenesis in cultured cells. (A) Subcellular distribution of SNX10. RCC10/VHL cells are transfected with plasmid-encoding SNX10-GFP or SNX10-Flag plus Rab11-GFP. Cells are fixed 24 h after transfection and stained with the indicated endogenous markers: PCM-1 (pericentriolar material), Ninein or Pericentrin (centrosome). A fraction of SNX10 is detected around Rab11 (recycling endosome), PCM-1, Ninein or Pericentrin. In ciliated cells, SNX10 localizes to the base of cilia. (B) SNX10 is required for ciliogenesis in RCC10/VHL cells. Cells are transfected with a scramble control siRNA (siCTL), a positive control siRNA (siRab8a) or siRNAs to SNX10. Cilia are induced by serum starvation and visualized by immunofluorescence staining with the anti-acetylated α-tubulin antibody (red). Nuclei are counterstained with DAPI (blue). The percentages of cells with cilia are reduced upon treatment with siRNAs to SNX10. (C) Real-time RT-PCR analysis for the efficiency of siRNAs to Rab8a and SNX10. β-actin is used as the internal control. The expression level of Rab8a or SNX10 in the siCTL-treated cells is arbitrary set as 100%. More than 80% inhibition of the target gene at the mRNA level is achieved in each case. (D) Statistical analysis for B. Assays are repeated at least three times and at least 400 cells are counted for each treatment. Data represent mean ± SD from three independent experiments (P < 0.002 for all three siRNAs to SNX10).
Figure 3
Figure 3
V-ATPase binds to SNX10 and mediates SNX10-induced vacuole formation. (A) V-ATPase is required for SNX10 to induce vacuoles in Hela cells. Pretreatment of cells with siRNA to the Vod1subunit of V-ATPase inhibits the formation of SNX10-GFP(10G)-induced vacuoles. Treatment of cells with the V-ATPase inhibitor bafilomycin A1 (baf) also inhibits the SNX10-GFP-induced vacuoles. (B) SNX10 co-localizes with V-ATPase. SNX10-GFP is co-transfected into RCC10/VHL cells with the HA-tagged Vod1 or V1D subunit of V-ATPase. Cells are fixed 24 h after transfection and immunostained with the mouse anti-HA antibody (red). Either subunit of the V-ATPase complex co-localizes with a fraction of SNX10-GFP. (C) SNX10 co-immunoprecipitates with the V1D subunit of V-ATPase. Left panel: HA-tagged SNX10 is co-transfected into 293T cells with the Flag-tagged GFP or Flag-GFP-tagged V1D. Cells are harvested 36 h after transfection and immunoprecipitation was performed with the resin-conjugated anti-Flag antibody. The levels of HA-SNX10 in cell lysates and immunoprecipitated complexes are detected by western blot using the mouse anti-HA IgG. V1D but not the Flag-GFP can pull down the HA-SNX10. Middle panel: Flag-GFP-tagged full-length SNX10 (10G) or the PX domain of SNX10 (PX, a.a. 8-125), but not the C-terminal domain of SNX10 (CD, a.a. 126-202), co-immunoprecipitates with HA-V1D. SNX16 (16G), another PX domain protein, does not pull down V1D. Right panel: HA-tagged V1A (A) and V1C (C) subunit of V-ATPase do not bind to SNX10. When they are co-expressed with V1D (D), SNX10 only pulls-down the V1D.
Figure 4
Figure 4
V-ATPase regulates ciliogenesis in vitro. (A) Subcellular distribution of V-ATPase in RCC10/VHL cells. Cells are transfected with the GFP-tagged Vod1 or V1D and immunostained for endogenous PCM-1, Pericentrin or Ninein 24 h after transfection. For ciliogenesis, cells are further cultured in serum-free RPMI-1640 media for 24 h then immunostained with the mouse anti-acetylated α-tubulin antibody (red). V-ATPase co-localizes with PCM-1, Ninein and Pericentrin and it is present at the base of cilia. (B) Real-time RT-PCR analysis for the efficiency of siRNAs to the V1D subunit of V-ATPase. They all induce more than 80% inhibition of V1D at the mRNA level. (C) V-ATPase is required for ciliogenesis in vitro. RCC10/VHL cells are transfected with a scramble control siRNA (siCTL) or siRNAs to V1D then serum starved to induce ciliogenesis. The percentages of cells with cilia are determined as described in Figure 2. Knockdown of V1D inhibits cilia formation (P < 0.01 for all three siRNAs). Data represent mean ± SD from three independent experiments.
Figure 5
Figure 5
V-ATPase regulates ciliogenesis in multiple organs in zebrafish. (A) Embryonic expression of Vod1 at day 1 as detected by in situ hybridization. It is widely distributed. (B) Knockdown of Vod1 by morpholino impairs the pigmentation in the trunk but not the retina. Similar defect is observed in V1D morphants (data not shown). (C) Morpholino knockdown of Vod1 causes heart-looping defect similar to that in SNX10a morphants. Furthermore, Vod1 and SNX10 function synergistically in the left-right patterning of heart. 10a-SP at 1.0 ng/embryo or Vod1-AUG at 0.75 ng/embryo alone cannot effectively induce heart-looping defect, however, the combination of them is sufficient to disrupt the heart-looping process. (D) Inhibition of V-ATPase induces the formation of pronephric cyst in 65% of the treated embryos (n = 46). Picture shows a cross-section of Vod1 morphant at the position of the cyst (arrows). NC: notochord. (E-G) V-ATPase is involved in ciliogenesis in multiple organs in zebrafish. Cilia in KV (E, 10-somite stage), pronephric duct (F, 27 hpf) or hair cells of lateral line neuromasts (red staining in G, day 2) are visualized by whole-mount fluorescence immunostaining. The nuclei are counter-stained with DAPI (blue). Ciliogenesis in the KV of V-ATPase morphants are severely inhibited. The cilia in pronephric duct are disorganized and reduced in number when V-ATPase is blocked. The neuromasts are relatively normal in the morphants (blue staining in G), but their hair cells fail to generate cilia.
Figure 6
Figure 6
SNX10 regulates the intracellular trafficking of V-ATPase and Rab8a. (A) SNX10 regulates the centrosomal localization of V-ATPase. RCC10/VHL cells are first treated with siRNAs to SNX10 for 24 h, plasmids for Vod1-GFP or V1D-GFP are then transfected into these cells. Cells are fixed and immunostained for Pericentrin after 24 h. In control siRNA-treated cells, a fraction of Vod1-GFP co-localizes with Pericentrin (red). In SNX10 siRNAs-treated cells, the centrosomal localization of Vod1-GFP is reduced. Vod1-GFP at other subcellular locations appears not affected. (B) Statistical analysis of A. (C) Western blot analysis for Vod1-GFP protein level with an anti-GFP antibody. The expression level of Vod1-GFP is not affected by the indicated siRNAs. Tubulin is the loading control. (D) Inhibition of SNX10 or V-ATPase does not affect the co-localization of Pericentrin and γ-tubulin. (E) Subcellular localization of SNX10 and Rab8a. Rab8a-GFP is enriched at the pericentriolar region and SNX10 vesicles are observed around this region. (F) SNX10 and V-ATPase regulate the ciliary trafficking of Rab8a. The Rab8a-GFP stable cell line is treated with the indicated siRNAs for 24 h then serum starved for another 24 h. Cells are then fixed for immunostainig. In the control siRNA-treated cells, 43% of the cilia is Rab8a-GFP positive. When SNX10 or V-ATPase is knocked down, in a subset of cells that form cilia, the ratio of Rab8a positive cilia is reduced. (G) Statistical analysis of F. The ratio of Rab8a positive cilia is reduced to less than 21% in all cases when SNX10 or V-ATPase is inhibited (P< 0.05 for all treatments when compared to the siCTL). (H) Inhibition of SNX10 or V-ATPase does not change the expression of Rab8a-GFP. The protein level of Rab8a-GFP is determined by western blot with an antibody to GFP and tubulin is used as the loading control.

Similar articles

See all similar articles

Cited by 23 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback