Assembly of centrosomal proteins and microtubule organization depends on PCM-1

Abstract

The protein PCM-1 localizes to cytoplasmic granules known as "centriolar satellites" that are partly enriched around the centrosome. We inhibited PCM-1 function using a variety of approaches: microinjection of antibodies into cultured cells, overexpression of a PCM-1 deletion mutant, and specific depletion of PCM-1 by siRNA. All approaches led to reduced targeting of centrin, pericentrin, and ninein to the centrosome. Similar effects were seen upon inhibition of dynactin by dynamitin, and after prolonged treatment of cells with the microtubule inhibitor nocodazole. Inhibition or depletion of PCM-1 function further disrupted the radial organization of microtubules without affecting microtubule nucleation. Loss of microtubule organization was also observed after centrin or ninein depletion. Our data suggest that PCM-1-containing centriolar satellites are involved in the microtubule- and dynactin-dependent recruitment of proteins to the centrosome, of which centrin and ninein are required for interphase microtubule organization.

Figure 1.

PCM-1 localizes to cytoplasmic granules that show a dynamic distribution during the cell cycle. (A) immunoblots of HeLa cell extracts and Xenopus egg extracts probed with rabbit preimmune serum, immune serum against PCM-1, and the same immune serum after affinity purification. Positions of molecular mass markers are indicated on the left. (B and C) HeLa cell stained with antibodies against (B) PCM-1, and against (C) γ-tubulin. (D) Live image of a HeLa cell transfected with full-length PCM-1, tagged with GFP at the carboxy terminus. (E–H) Double immunofluorescence of PCM-1 (green) and tubulin (red), and staining of the DNA (blue) of HeLa cells (E) in interphase, (F) prophase, (G) metaphase, and (H) telophase. PCM-1 signal in E–H was photographed at identical exposure levels. Bars: (D, E, G, and H) 10 μm; same magnifications in B–D and F and G, respectively.

Figure 2.

Microinjection of antibodies against PCM-1 causes aggregation of centrin and pericentrin. Xenopus A6 cells were microinjected with affinity-purified antibodies against PCM-1 (B, D–F, H, and J), or with control antibodies (A, C, G, and I). Cells were stained for immunofluorescence of (A and B) PCM-1, (C–F) centrin, (G and H) pericentrin, or (I and J) γ-tubulin. Insets show the same cells stained with Texas red–labeled anti–rabbit antibody to identify microinjected cells, and stained with DAPI to detect DNA (blue). Bars: (B and J) 10 μm; same magnifications in A and B and C–J, respectively.

Figure 3.

Overexpression of the PCM-1 deletion mutant 1–1468 delocalizes centrin, pericentrin, and endogenous PCM-1 to large cytoplasmic aggregates. (A and B) Control HeLa cells overexpressing β-galactosidase, stained in red; DNA stained in blue. (B) PCM-1 is stained in green. (C, E, G, and M) HeLa cells overexpressing PCM-1 deletion mutant 1–1468, stained in red; DNA stained in blue. The same cells were stained for (D) PCM-1, (F) centrin, (H) pericentrin, or (N) acetylated tubulin. The arrowhead in N indicates the position of the centriole pair. (I–L) CHO cells overexpressing PCM-1 mutant 1–1468 (I and K; red), stained for (J) γ-tubulin, or (L) ninein. Bars: (H and N) 10 μm; same magnifications in A, B, E–H and C, D, I–N, respectively.

Figure 4.

Depletion of PCM-1 by RNA silencing reduces centrosomal localization of centrin, pericentrin, and ninein, but not γ-tubulin or dynactin. A–J show U-2 OS cells treated with control or PCM-1 siRNA oligonucleotides, as indicated. Image pairs show cells double stained for (A and B) PCM-1 and centrin, (C and D) PCM-1 and pericentrin, (E and F) PCM-1 and ninein, (G and H) PCM-1 and γ-tubulin, and (I and J) PCM-1 and dynactin p150/glued. The amount of centrosomal protein localization after PCM-1 depletion was determined by photometric analysis to be 39% of centrin (± 17), 36% of pericentrin (± 21), 38% of ninein (± 20), 99% of γ-tubulin (± 58), and 82% of dynactin (± 37), as compared with control cells (n = 34 cells/each). (K) Immunoblots of extracts from untreated cells (untr.), and cells treated with control RNA oligomers siRNA PCM-1.1 or siRNA PCM-1.2, for different lengths of time as indicated. Blots were probed with antibodies against PCM-1, NuMA ,and centrin-3. Bar (J), 10 μm.

Figure 5.

PCM-1 binds to centrin, and partly colocalizes with centriolar satellites of centrin, ninein, and pericentrin. (A) Binding assay using glutathione beads on cell extracts preincubated with GST-tagged centrin-3 or with GST alone. Lanes on a Coomassie-stained gel show relative molecular mass markers (Mr), purified GST-centrin-3, the eluate from glutathione beads incubated with Xenopus egg extract and GST-centrin-3, egg extract alone, the eluate from glutathione beads incubated with egg extract and GST, and purified GST alone. Positions of molecular weight markers are indicated on the left. Shown below are immunoblots of corresponding lanes, probed with antibodies against PCM-1. (B) Immunofluorescence staining of centriolar satellites. Left column, confocal section of a PtK₂ cell stained for centrin-3 (green) and PCM-1 (red). Middle and right columns, conventional immunofluorescence of mouse myoblast cells stained for (middle) ninein and PCM-1, or (right) pericentrin and PCM-1. Bar (B), 10 μm.

Figure 6.

Microtubules and dynactin are essential for the centrosomal accumulation of PCM-1, centrin, pericentrin, and ninein, but not γ-tubulin. (A–J) Image pairs of CHO cells after nocodazole treatment are shown, stained for (A and B) PCM-1 and centrin, (C and D) PCM-1 and pericentrin, (E and F) PCM-1 and ninein, (G and H) PCM-1 and γ-tubulin, and (I and J) PCM-1 and dynein intermediate chain. K and L show an untreated cell stained for PCM-1 and dynein intermediate chain. Arrowheads indicate pericentriolar dynein spots colocalizing with PCM-1 granules. (M–V) Image pairs of CHO cells microinjected with p50/dynamitin are shown. (M, O, Q, S, and U) Cells were stained for PCM-1, centrin-3, ninein, γ-tubulin, and pericentrin, respectively. (N, P, R, T, and V) Corresponding images showing dynamitin-injected cells (red) and DNA staining (blue). Dynamitin-dependent inhibition of centrosomal localization varied for different proteins; PCM-1 dispersed in 77% of injected cells (n = 106, controls 2%, n = 182), centrin was affected in 60% (n = 50, controls 2%, n = 191), ninein in 45% (n = 110, controls 3%, n = 169), pericentrin in 33% (n = 470, controls 5%, n = 73),and γ-tubulin in 3% (n = 76, controls 1%, n = 135). (W and X) Image pair of a control cell injected with labeled goat anti–rabbit antibody and stained for pericentrin. Bar (L), 10 μm.

Figure 7.

Microtubule anchoring to the centrosome depends on PCM-1. (A and B) Cos-7 cells expressing GFP-tagged PCM-1 deletion mutant 1–1468 (A, green). (A) Microtubules are stained in red, B shows microtubules only. (C and D) Control cells overexpressing GFP only (C, green). (C) Microtubules in red, (D) microtubules only. (E–J) Microtubule regrowth after nocodazole treatment of PtK₂ cells. (E, G, and I) Staining of microtubules in red, GFP-tagged PCM-1 mutant 1–1468 in green; (F, H, and J) microtubules only. Time points at (E and F) 0 min, (G and H) 5 min, and (I and J) 60 min after removal of the drug. (K) Graph showing percentage of cells with radial microtubule organization at different time points after removal of nocodazole in untransfected cells (blue), cells overexpressing GFP (green), and cells overexpressing GFP-tagged PCM-1 mutant 1–1468 (red). Between 400 and 600 cells were counted for each time point. Bars: 10 μm (D and J).

Figure 8.

Depletion of PCM-1, centrin-3, or ninein results in loss of microtubule anchoring at the centrosome. (A C, E, and G) U-2 OS cells were transfected with control dsRNA oligomers and stained in green for (A) PCM-1, (C) centrin-3, (E) ninein, and (G) pericentrin. Microtubules were stained in red, DNA in blue. (B, D, F, and H) Corresponding image pairs showing cells after treatment with siRNA against (B) PCM-1, (D) centrin-3, (F) ninein, (H) pericentrin. I, J, U-2 OS cells treated with (I) control oligomers or (J) centrin-3 siRNA. Red, centrin-3 immunofluorescence; green, ninein immunofluorescence. K, immunoblots of cells treated with control RNA or siRNA against (left) centrin-3, or (right) ninein, stained for centrin-3 or ninein, respectively. Bar in H, 10 μm.