A ciliopathy complex builds distal appendages to initiate ciliogenesis

Abstract

Cells inherit two centrioles, the older of which is uniquely capable of generating a cilium. Using proteomics and superresolved imaging, we identify a module that we term DISCO (distal centriole complex). The DISCO components CEP90, MNR, and OFD1 underlie human ciliopathies. This complex localizes to both distal centrioles and centriolar satellites, proteinaceous granules surrounding centrioles. Cells and mice lacking CEP90 or MNR do not generate cilia, fail to assemble distal appendages, and do not transduce Hedgehog signals. Disrupting the satellite pools does not affect distal appendage assembly, indicating that it is the centriolar populations of MNR and CEP90 that are critical for ciliogenesis. CEP90 recruits the most proximal known distal appendage component, CEP83, to root distal appendage formation, an early step in ciliogenesis. In addition, MNR, but not CEP90, restricts centriolar length by recruiting OFD1. We conclude that DISCO acts at the distal centriole to support ciliogenesis by restraining centriole length and assembling distal appendages, defects in which cause human ciliopathies.

Figure 1.

CEP90 localizes to centriolar satellites and the distal centriole. (a) 3D-SIM of immunostained RPE1 cells reveals localization of CEP90 (yellow) at centriolar satellites (PCM1, cyan) and centrioles (Tubulin^Ac, magenta) in WT, cycling RPE1 cells. Scale bar = 1 µm. (b) Treatment of serum-starved RPE1 cells with nocodazole disperses the centriolar satellites, highlighting 3D-SIM of CEP90 (yellow) rings at centrioles (γ-tubulin, magenta). Distal appendage component CEP164 (cyan) indicates the mother centriole in c and d. Scale bar = 1 µm. (c) 3D-SIM of serum-starved PCM1^−/− RPE1 cells shows that CEP90 (yellow) localizes to centrioles (γ-tubulin, magenta) independent of centriolar satellites. Scale bar of main panel and insets = 1 µm and 0.5 µm respectively. (d) 3D-SIM confirms localization of CEP90 (yellow) at centrioles (γ-tubulin, magenta) in nocodazole-treated, serum-starved eYFP-CEP90–expressing PCM1^−/− RPE1 cells. Scale bar = 1 µm. (e) Quantification of eYFP-CEP90, Centrobin, and CEP164 fluorescence intensity at daughter (DC) and mother (MC) centrioles from 3D-SIM images, n = 10–20 cells. Horizontal lines indicate means ± SEM. ***, P < 0.0005, unpaired t test. AU, arbitrary units. (f) LR-ExSIM of RPE1 cells immunostained for CEP90 (yellow) and CEP164 (magenta) reveals that rings of CEP90 are composed of discrete puncta. CEP90 rings are smaller and more proximal to CEP164 rings. Scale bar = 1 µm. (g) Example of an LR-ExSIM image of a radially oriented centriole used to quantify the number and angle between adjacent puncta of CEP90. Scale bar = 0.5 µm. (h) Histogram of number of discrete puncta of CEP90 and CEP164 observed per centriole in LR-ExSIM images. n = 12–17 measurements. (i) Histogram of the angular spacing between adjoining centriolar CEP90 puncta observed by LR-ExSIM. n = 66 measurements. (j) Schematic of the ring of CEP90 punctae (yellow), distal appendages (magenta), and subdistal appendages (blue) at the distal centriole. CEP90 decorates the distal end of mother and daughter centrioles.

Figure S1.

Generation of CEP90^−/−, MNR^−/−, and PCM1^−/− cell lines.(a–c) Sequence analysis of genomic DNA isolated from control and CEP90^−/− (a), MNR^−/− (b), and PCM1^−/− (c) RPE1 cell lines generated using CRISPR-Cas9. Insertions and deletions, and translation products resulting from genome editing are indicated. The protospacer adjacent motif (PAM) is indicated in green. (d) Immunoblot analysis of whole-cell lysates from control, CEP90^−/−, MNR^−/−, and PCM1^−/− RPE1 cell lines confirms loss of protein in mutant cell lines. The specific MNR band is indicated with an asterisk, and the top band is nonspecific.

Figure 2.

CEP90 forms a complex with OFD1 and MNR. (a) Schematic depicting workflow used to identify CEP90 interactors at centriolar satellites and centrioles using WT and PCM1^−/− RPE1 cells. PCM1^−/− cells localize CEP90 to the distal centriole, and WT cells localize CEP90 to the distal centriole and centriolar satellites. (b) Venn diagram comparing high-confidence interactors of CEP90 in WT and PCM1^−/− RPE1 cells. (c) Interactome dot representation of selected CEP90 interactors identified in the proteomic screen. Hits were grouped based on their cellular localization. Average number of peptide spectra (AvgSpec) is represented by dot shade. Abundance of a peptide spectrum produced in relation to the most abundant spectrum is depicted by dot size. Bayesian false discovery rate (BFDR) is represented by rim color. (d) Immunoblot of a subset of CEP90 interactions identified by proteomics were validated by coimmunoprecipitation (IP). CEP90 interacts with PCM1, OFD1 and MNR, but not α-tubulin or GAPDH. FT, flow through. Specific MNR band is indicated with an asterisk. The top band is nonspecific, as it is undiminished in the MNR-knockout cell lysates. (e) Coulson plot showing the phylogenetic distribution of a subset of centriolar proteins in select ciliated metazoan species. Orthologues identified with high confidence are indicated with a filled circle, and a subset of CEP90 interactors further explored in this study are highlighted in blue. The dendrogram on top (made using interactive tree of life; Ciccarelli et al., 2006) shows the evolutionary relationship between species.

Figure 3.

CEP90 colocalizes with OFD1 and MNR. (a) Immunostaining of RPE1 cells for CEP90 (yellow), MNR (cyan), and γ-tubulin (magenta) demonstrating that CEP90 colocalizes with MNR at centriolar satellites. Scale bar = 2 µm. (b) 3D-SIM of RPE1 cells treated with nocodazole to disperse centriolar satellites highlights colocalization of CEP90 (yellow) and MNR (cyan) at centrioles (γ-tubulin, magenta). Scale bar = 1 µm. (c) Immunostaining of OFD1 (yellow), MNR (cyan), and γ-tubulin (magenta) reveals CEP90 and MNR colocalization at centriolar satellites. Scale bar = 2 µm. (d) 3D-SIM of immunostained cells treated with nocodazole reveals a ring of OFD1 (yellow) colocalizing with MNR (cyan) at centrioles (γ-tubulin, magenta). Scale bar = 1 µm. (e) Measurements of ring diameters measured in 3D-SIM images. For distal centriole proteins, ring diameters were measured at the mother (MC) and daughter centriole (DC). Scatter dot plot shows mean ± SD. *, P < 0.05, unpaired t test. n = 13–22 centrioles. (f) Schematic representation of a radial view of a mother centriole with distal appendages (purple), OFD1 (orange), CEP90 (yellow), and MNR (green). (g) Localization of distal centriole proteins (yellow) to centrioles (γ-tubulin, magenta) is cell cycle dependent. Cells in G1 phase were identified as lacking both EdU and CENPF staining, cells in S phase as being positive for EdU but lacking CENPF, and cells in G2 phase as being positive for CENPF but lacking EdU staining. Two puncta of CEP90, OFD1, and MNR were observed at one centrosome during G1 and S phases and four puncta at two centrosomes after centrosome duplication during G2 phase. Scale bars represent 5 µm in main panels and 0.5 µm in insets.

Figure 4.

CEP90 and MNR are essential for ciliogenesis. (a) WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells were immunostained for cilia (ARL13B, magenta), centrosomes (γ-tubulin, yellow), and nuclei (Hoechst, blue). Scale bar = 10 µm. (b) Quantification of ciliation frequency of WT, CEP90^−/−, and MNR^−/− RPE1 cells serum starved for times indicated. n > 100 cells from two biological replicates. Bar graph shows mean ± SEM. (c) Images of control, Cep90^−/−, and Mnr^−/− embryos at E9.5. Scale bar = 100 µm. (d) Whole-mount immunostaining of nodes of littermate control and Cep90^−/− embryos at E8.5 for centrosomes (γ-tubulin, yellow), cilia (Tubulin^Ac, magenta), and nuclei (Hoechst, blue). Arrows in the control image point to cilia projecting into the node lumen, which are absent in Cep90^−/− embryos. Scale bars represent 5 µm in the main panels and 2 µm in insets. (e) In situ hybridization for Gli1 in E8.5 littermate control and Cep90^−/− embryos revealing decreased expression in the absence of CEP90 indicative of disrupted Hedgehog signaling. Scale bar = 300 µm. (f) Immunostaining of E9.5 neural tube sections of littermate control and Mnr^−/− embryos for centrosomes (FOP, yellow), cilia (ARL13B, magenta), and nuclei (Hoechst, blue), indicating that MNR is required for ciliogenesis in vivo. Scale bar = 5 µm. (g) MEFs derived from Cep90^+/+ and Cep90^−/− embryos were serum starved for 24 h and immunostained for cilia (ARL13B, magenta), centrosomes (γ-tubulin, yellow), and nuclei (Hoechst, blue). Scale bar = 10 µm. (h) Quantification of ciliation frequency shows loss of cilia in Cep90^−/− MEFs. Bar graph shows mean ± SEM. Asterisks indicate P < 0.0005 determined using unpaired t test. n > 100 cells from two biological replicates. (i) qRT-PCR of HH target genes Gli1 and Ptch1 in serum-starved Cep90^+/+ and Cep90^−/− MEFs stimulated with 200nM SAG for 24h relative to DMSO-treated controls. Bar graph shows mean ± SEM. Asterisks indicate P < 0.05 determined using unpaired t test (**, P < 0.005; ***, P < 0.0005). n = 3 biological replicates. (j) MEFs derived from Mnr^+/+ and Mnr^−/− embryos were serum starved for 24 h and immunostained for cilia (ARL13B, magenta), centrosomes (γ-tubulin, yellow) and nuclei (Hoechst, blue). Scale bar = 10 µm. (k) Quantification of ciliation frequency shows loss of cilia in Mnr^−/− MEFs. Bar graph shows mean ± SEM. Asterisks indicate P < 0.0005 determined using an unpaired t test. n > 100 cells from two biological replicates. (l) qRT-PCR of HH target genes Gli1 and Ptch1 in serum-starved Mnr^+/+ and Mnr^−/− MEFs stimulated with 200 nM SAG for 24 h relative to DMSO-treated controls. Bar graph shows mean ± SEM. Asterisks indicate P < 0.05 determined using unpaired t test (***, P < 0.0005; *, P < 0.05). n = 3 biological replicates.

Figure 5.

CEP90 ciliopathy mutations affect ciliogenesis and centriolar satellite morphology. (a) Schematic representation of full-length (FL) human CEP90, truncation constructs, and disease-associated mutations. Coiled-coil (CC) domains identified by MARCOIL (Zimmermann et al., 2018) at 90% threshold are indicated in cyan. (b) Quantification of ciliation frequency of CEP90^−/− RPE1 (control) and CEP90^−/− cells expressing mNeonGreen-tagged disease variants and truncations of CEP90. Graph shows mean ± SEM. ***, P < 0.05, one-way ANOVA. n > 100 cells from two biological replicates. (c) CEP90^−/− RPE1 cells and CEP90^−/− cells expressing the denoted mNeonGreen (mNG)–tagged versions of CEP90 were serum starved for 24 h and immunostained for cilia (ARL13B), centriolar satellites (CEP131), and mNeonGreen. Scale bar = 2 µm. (d) Immunoprecipitation of GFP-tagged full-length and truncation constructs of CEP90 blotted for GFP and OFD1. (e) Quantification of OFD1 band intensities relative to corresponding GFP input band intensities. (f) Immunoprecipitation of GFP-tagged full-length and truncation constructs of CEP90 blotted for GFP and MNR. (g) Quantification of MNR band intensities relative to corresponding GFP input band intensities. Asterisk represents the MNR band; the top band is nonspecific.

Figure 6.

MNR, but not CEP90, restricts centriole length. (a) 3D-SIM images of WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for cilia/centrioles (Tubulin^Ac, magenta) and CEP162 (yellow), a distal centriolar protein. Scale bars represent 1 µm in the main panels and 0.5 µm in insets. (b) Graph of centriolar lengths measured using 3D-SIM images. Centriole lengths are longer and have a wider distribution in MNR^−/− cells than in WT or CEP90^−/− cells. Error bars indicate mean ± SEM. ***, P < 0.05, one-way ANOVA. n = 14–30 centrioles. (c) Histogram of centriole lengths observed in WT, MNR^−/−, and CEP90^−/− cells. n = 14–30 centrioles. (d) Serial-section TEM confirms the presence of elongated centrioles in MNR^−/− RPE1 cells. Scale bar = 500 nm. (e) Centriole lengths of WT and MNR^−/− RPE1 cells measured using TEM images. Horizontal lines indicate means ± SEM. Asterisks indicate P < 0.005 determined using unpaired t test. n = 8 centrioles per condition. (f) WT and MNR^−/− RPE1 cells were serum-starved and immunostained with antibodies to Tubulin^Ac and subdistal appendage component Ninein to distinguish mother (MC) and daughter centrioles (DC). Graph of centriolar lengths measured using 3D-SIM. MNR restrains centriole lengthening of both mother and daughter centrioles. Error bars indicate mean ± SEM. Asterisks indicate P < 0.05 determined using unpaired t test (*, P < 0.05; **, P < 0.005). n = 10–35 measurements. (g) Schematic depicting distinct roles of CEP90 and MNR in regulating centriole length. (h) 3D-SIM imaging of serum-starved WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for OFD1 (yellow), centrioles (γ-tubulin, cyan), and cilia (Tubulin^Ac, magenta). Boxed regions are depicted in insets throughout. OFD1 localizes to centrioles in WT and CEP90^−/− cells, but not MNR^−/− cells. Scale bar = 1 µm. (i) Quantification of OFD1 fluorescence intensity at centrioles in WT, CEP90^−/−, and MNR^−/− cells. Horizontal lines indicate means ± SEM. ***, P < 0.05, one-way ANOVA. n = 64–187 cells.

Figure S2.

Overexpressed MNR localizes to microtubules and recruits endogenous OFD1. (a) RPE1 cells transiently transfected with MYC-tagged MNR and immunostained with antibodies to MYC-tag and α-tubulin. Scale bar = 10 µm. (b) RPE1 cells transiently transfected with MYC-tagged MNR and immunostained with antibodies to MYC-tag and OFD1. Scale bar = 10 µm. (c) RPE1 cells transiently transfected with MYC-tagged MNR and immunostained with antibodies to MYC-tag and CEP90. Scale bar = 10 µm. (d) RPE1 cells transiently transfected with MYC-tagged MNR and immunostained with antibodies to MYC-tag and FOPNL. Scale bar = 10 µm.

Figure 7.

CEP90 and MNR are required for early ciliogenesis. (a) WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for IFT88 (yellow), centrioles (γ-tubulin, cyan), and cilia (Tubulin^Ac, magenta. IFT88 is not recruited to the centrosome of CEP90^−/− or MNR^−/− cells. (b) Quantification of IFT88 fluorescence intensity at WT, CEP90^−/−, and MNR^−/− centrosomes. Error bars indicate mean ± SEM. ***, P < 0.05, unpaired t test. n = 38–63 measurements. (c) WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for CP110 (yellow), centrioles (γ-tubulin, cyan), and cilia (Tubulin^Ac, magenta). Scale bar = 1 µm. (d) Quantification of whether CP110 localizes to one or two centrioles. In the absence of CEP90 or MNR, CP110 continues to localize to the distal mother centriole. n > 50 cells from two independent experiments. Error bars indicate mean ± SEM. (e) WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for CEP97 (yellow), γ-tubulin (cyan), and Tubulin^Ac (magenta). Scale bar = 1 µm. (f) Quantification of whether CEP97 localizes to one or two centrioles. As with CP110, CEP90 and MNR are required to remove CEP97 from the distal mother centriole. n > 50 cells from two independent experiments. (g) 3D-SIM images of RPE1 cells immunostained for Myosin Va (Myo-Va, yellow) and centrioles (γ-tubulin, magenta). Myo-Va cannot localize near centrosomes (left), can localize to preciliary vesicles (denoted centrosomal Myo-Va, middle), or can localize to the ciliary pocket (denoted ciliary Myo-Va, right). Scale bar = 1 µm. (h) Quantification of three distinct Myo-Va staining patterns in WT, CEP90^−/− and MNR^−/− cells. n > 50 cells from two independent experiments. (i) Serial-section TEM images of serum-starved Cep90^+/+ and Cep90^−/− MEFs confirms the absence of preciliary vesicle docking at the Cep90^−/− mother centriole. Scale bar = 200 nm. n = 9 cells for both genotypes. (j) Serial-section TEM images of WT and MNR^−/− RPE1 cells confirms the absence of preciliary vesicle docking at the MNR^−/− mother centriole. Scale bar = 250 nm. n = 10 cells for WT and n = 6 for MNR^−/− cells. (k) Model based on our data highlighting the role of CEP90 and MNR in maturation of the mother centriole.

Figure S3.

CEP90 and MNR regulate distal appendage assembly irrespective of ciliation status. (a) Cycling WT, CEP90^−/−, and MNR^−/− RPE1 cells stained with antibodies to IFT88, γ-tubulin (centrosome marker), and Tubulin^Ac (cilia and centriole marker). 3D-SIM imaging reveals IFT88 at the centrosome in WT, but not CEP90^−/− and MNR^−/−, cells. Scale bars for main panels and insets represent 1 µm. (b) Quantification of IFT88 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ***, P < 0.001, one-way ANOVA. n = 73–96 measurements. (c) Cycling WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for CP110 (yellow), centrioles (FOP, cyan), and Tubulin^Ac (cilia and centriole marker, magenta). Scale bar = 1 µm. (d) Quantification of CP110 foci. (e) Cycling WT, CEP90^−/−, and MNR^−/− RPE1 cells stained with antibodies to FBF1, γ-tubulin (centrosome marker), and Tubulin^Ac (cilia and centriole marker). 3D-SIM imaging reveals FBF1 at the mother centriole in WT, but not CEP90^−/− and MNR^−/−, cells. Scale bars for main panels and insets represent 1 µm. (f) Quantification of FBF1 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ***, P < 0.001, one-way ANOVA. n = 21–45 measurements. (g) Cycling WT, CEP90^−/−, and MNR^−/− RPE1 cells stained with antibodies to CEP164, γ-tubulin (centrosome marker), and Tubulin^Ac (cilia and centriole marker). 3D-SIM imaging reveals CEP164 at one of the two centrioles in WT, but not CEP90^−/− and MNR^−/−, cells. Scale bars for main panels and insets represent 1 µm. (h) Quantification of CEP164 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ***, P < 0.001, one-way ANOVA. n = 37–53 measurements.

Figure S4.

Serial-section TEM confirms loss of distal appendages in cells lacking CEP90 or MNR. (a) TEM images ofCep90^+/+ and Cep90^−/− MEF cells confirms the absence of preciliary vesicle docking and distal appendages at the Cep90^−/− mother centriole. Blue arrowheads indicate subdistal appendages, and pink arrowheads indicate distal appendages. (b) TEM images of WT and MNR^−/− RPE1 cells confirms the absence of preciliary vesicle docking and distal appendages at the MNR^−/− mother centriole. Centrioles are hyperelongated in the absence of MNR. Blue arrowheads indicate subdistal appendages, and pink arrowheads indicate distal appendages. Scale bars = 500 nm. CP, ciliary pocket.

Figure S5.

CEP90 regulates distal appendage assembly independent of Talpid3 recruitment and removal of Centrobin.(a) 3D-SIM imaging of serum-starved WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for Ninein (yellow), centrioles (γ-tubulin, cyan), and cilia (ARL13B, magenta). Boxed regions are depicted in insets throughout. Ninein localizes to centrioles in WT, CEP90^−/−, and MNR^−/− cells. Scale bar = 1 µm. (b) Quantification of Ninein fluorescence intensity at centrioles in WT, CEP90^−/−, and MNR^−/− cells. Horizontal lines indicate means ± SEM. Asterisks indicate P < 0.05 determined using one-way ANOVA. n = 36–45 measurements. (c) 3D-SIM imaging of serum-starved WT, CEP90^−/− and MNR^−/− RPE1 cells immunostained for CEP170 (yellow), centrioles (γ-tubulin, cyan) and cilia (Tubulin^Ac, magenta). Boxed regions are depicted in insets throughout. CEP170 localizes to centrioles in WT, CEP90^−/−, and MNR^−/− cells. Scale bar = 1 µm. (d) Quantification of CEP170 fluorescence intensity at centrioles in WT, CEP90^−/−, and MNR^−/− cells. Horizontal lines indicate means ± SEM. Asterisks indicate P < 0.005 determined using one-way ANOVA. n = 73–99 measurements. (e) Immunoblot of distal appendage proteins CEP83 and CEP164 in WT, CEP90^−/−, MNR^−/−, and PCM1^−/− RPE1 cells. (f) Coimmunoprecipitation of daughter centriole proteins Centrobin and CEP120 with CEP90. IP indicates eluate, and FT indicates flow-through. (g) WT, CEP90^−/−, and MNR^−/− RPE1 cells were serum starved for 24 h and stained with antibodies to Talpid3, γ-tubulin (centrosome marker), and ARL13B (cilia marker). 3D-SIM imaging reveals ring of Talpid3 at mother and daughter centrioles. Scale bars represent 1 µm for main panels and insets. (h) Quantification of Talpid3 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. Asterisks indicate P < 0.0005, determined using one-way ANOVA. n = 33–35 measurements. (i) WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for Centrobin (yellow), centrioles (FOP, cyan), and distal appendages (CEP164, magenta). Scale bar = 1 µm. (j) Quantification of whether Centrobin localizes to one or two centrioles. n > 50 cells from two biological replicates. CEP90 and MNR are not required to remove Centrobin from the distal mother centriole. (k) WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for CEP120 (yellow), centrioles (FOP, cyan), and distal appendages (CEP164, magenta). Scale bar = 1 µm. (l) Quantification of whether CEP120 localizes to one or two centrioles. n > 50 cells from two biological replicates. CEP90 and MNR are not required to remove Centrobin or CEP120 from the distal mother centriole.

Figure 8.

CEP90 and MNR recruit distal appendage components to the mother centriole.(a–j) 3D-SIM images and quantification of centrosomal intensity of WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained for γ-tubulin (cyan), Tubulin^Ac (magenta), and distal centriole components (yellow) CEP83 (a and b), FBF1 (c and d), SCLT1 (e and f), ANKRD26 (g and h), and CEP164 (i and j). Scale bars = 1 µm. Horizontal lines in scatter dot plots indicate means ± SEM. ***, P < 0.001, one-way ANOVA. n ≥ 40 measurements per condition.

Figure 9.

Centriolar satellites are dispensable for distal appendage assembly at the mother centriole. (a) 3D-SIM images of WT, CEP90^−/−, and MNR^−/− RPE1 cells immunostained with antibodies to PCM1, a centriolar satellite marker, ARL13B and γ-tubulin. Scale bar = 2 µm. (b) Quantification of PCM1 intensity in the pericentrosomal area. PCM1-positive centriolar satellites fail to accumulate around centrosomes in CEP90^−/−, but not MNR^−/−, cells. ***, P < 0.05, ordinary one-way ANOVA. n = 37–44 measurements. (c) Ciliogenesis is disrupted in PCM1^−/− RPE1 cells 24 h after serum starvation. n > 100 cells from two biological replicates. Bar graph shows mean ± SEM. (d) WT and PCM1^−/− RPE1 cells were serum starved for 24 h and stained with antibodies to CEP83, γ-tubulin (centrosome marker), and Tubulin^Ac (cilia marker). 3D-SIM imaging reveals ring of CEP83 at the mother centrioles in PCM1^−/− RPE1 cells. Scale bar = 1 µm. (e) Quantification of CEP83 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ns, unpaired t test. n = 53–124 measurements. (f) WT and PCM1^−/− RPE1 cells were serum starved for 24 h and stained with antibodies to SCLT1, γ-tubulin (centrosome marker), and ARL13B (cilia marker). 3D-SIM imaging reveals ring of SCLT1 at the mother centrioles in WT and PCM1^−/− RPE1 cells. Scale bar = 1 µm. (g) Quantification of SCLT1 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ns, unpaired t test. n = 26–28 measurements. (h) WT and PCM1^−/− RPE1 cells were serum starved for 24 h and stained with antibodies to CEP164 and γ-tubulin (centrosome marker). 3D-SIM imaging reveals ring of CEP164 at the mother centrioles in PCM1^−/− RPE1 cells. Scale bar = 1 µm. (i) Quantification of CEP164 fluorescence intensity at centrioles. Scatter dot plots show mean ± SEM. ***, P < 0.001, unpaired t test. n = 40–60 measurements. (j) Representative TEM images of WT and PCM1^−/− RPE1 cells serum starved for 1 h. Distal appendages are marked with arrows. Scale bar = 200 nm.

Figure 10.

MNR recruits CEP90, which recruits CEP83 to build distal appendages. (a) 3D-SIM imaging of WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for CEP90 (yellow), γ-tubulin (cyan), and Tubulin^Ac (magenta). Scale bar = 1 µm. (b) Quantification of CEP90 fluorescence intensity at centrioles. Horizontal lines in scatter dot plots indicate means ± SEM. ***, P < 0.0001, one-way ANOVA. n = 31–37 measurements. CEP90 fails to localize to MNR^−/− and CEP90^−/− centrioles, although protein levels of CEP90 remained unchanged in MNR^−/− RPE1 cells (Fig. S1 d). (c) 3D-SIM imaging of WT, CEP90^−/−, and MNR^−/− serum-starved RPE1 cells immunostained for MNR (yellow), γ-tubulin (cyan), and CEP164 (magenta). Scale bar = 1 µm. (d) Quantification of MNR fluorescence intensity at centrioles. Horizontal lines in scatter dot plots indicate means ± SEM. ***, P < 0.0001, one-way ANOVA. n = 73–122 measurements. MNR localization is reduced but present at CEP90^−/− centrioles. (e) Quantification of Talpid3 diameter at mother (MC) and daughter centrioles (DC) in serum-starved WT and CEP90^−/− cells. Horizontal lines in scatter dot plots indicate means ± SEM. ***, P < 0.05, one-way ANOVA. n = 11–16 measurements. The Talpid3 ring diameter is increased at WT mother centrioles, but not CEP90^−/− mother centrioles. (f) Quantification of OFD1 diameter at mother (MC) and daughter centrioles (DC) in serum-starved WT and CEP90^−/− cells. Horizontal lines in scatter dot plots indicate means ± SEM. Asterisks indicate P < 0.05 (*, P < 0.05; **, P < 0.005; ***, P < 0.0005), one-way ANOVA. n = 17–20 measurements. In the absence of CEP90, the mother centriole ring of OFD1 does not expand as in WT cells. (g) WT or RPE1 cells expressing CEP83-HA were lysed, immunoprecipitated with anti-HA antibody–bound beads, and immunoblotted with antibodies to HA and CEP90. (h) Model of the hierarchical recruitment of the DISCO complex. MNR recruitment of OFD1 restrains centriole elongation and MNR recruits CEP90, which, in turn, recruits CEP83, the base of the distal appendage.