Self-organization of Plk4 regulates symmetry breaking in centriole duplication

Abstract

During centriole duplication, a single daughter centriole is formed next to the mother centriole. The molecular mechanism that determines a single duplication site remains a long-standing question. Here, we show that intrinsic self-organization of Plk4 is implicated in symmetry breaking in the process of centriole duplication. We demonstrate that Plk4 has an ability to phase-separate into condensates via an intrinsically disordered linker and that the condensation properties of Plk4 are regulated by autophosphorylation. Consistently, the dissociation dynamics of centriolar Plk4 are controlled by autophosphorylation. We further found that autophosphorylated Plk4 is already distributed as a single focus around the mother centriole before the initiation of procentriole formation, and is subsequently targeted for STIL-HsSAS6 loading. Perturbation of Plk4 self-organization affects the asymmetry of centriolar Plk4 distribution and proper centriole duplication. Overall, we propose that the spatial pattern formation of Plk4 is a determinant of a single duplication site per mother centriole.

Fig. 1

Plk4 modulates its solubility by autophosphorylation in vitro. a Disorder prediction and schematic of human Plk4 domains. Kinase kinase domain, L1/L2 Linker 1/2, LCR low complexity region, CPB Cryptic Polo-box, PB Polo-box. b Image of 500 nM Plk4 (Kinase + L1)-His₆ fragments after GST-tag cleavage by PreScission protease. This truncated form was used for in vitro experiments to directly examine the function of L1 in Plk4 condensation. c Amino acid sequence of low complexity region (288–303 a.a.) and its neighboring region of human Plk4. Red letters, mutation sites; Gray background, degron motif. d Spin-down assay. After GST-tag cleavage, 500 nM Plk4 (Kinase + L1)-His₆ solution was centrifuged and separated into supernatant (S) and pellet (P) fractions. Representative gel stained with CBB and the quantification are shown. Graph represents mean percentages ± SD of the pellet fraction from three independent experiments. NaCl, 500 mM. e Measurement of the light scattering of GST-Plk4 (Kinase + L1)-His₆ (100 μg/ml) under thermal control. Left: Representative data of three independent experiments. Values were normalized to the scattering at 25 °C. Dotted vertical lines indicate each aggregation onset temperature. Right: Aggregation onset temperature. Graph shows mean temperature ± SD from three independent experiments. f Measurement of protein aggregation of 300 nM GST-Plk4 (Kinase + L1)-His₆ by Proteostat aggregation assay. Fluorescence values were normalized to the WT. Graph shows mean ± SD from three independent experiments. g Dephosphorylation assay. GST-His₆ and GST-Plk4(Kinase + L1)-His₆ (100 μg/ml) were incubated with λPP at 30 °C for the indicated time and each light scattering was measured. Representative data of three independent experiments. h Schematic of in vitro solubility of Plk4. Source data are provided as a Source Data file

Fig. 2

Condensation properties of Plk4 are regulated by autophosphorylation. a Representative images of 70 nM GFP-Plk4 (Kinase) and GFP-Plk4 (Kinase + L1) in the presence of PEG. Scale bar, 2 μm. b Representative images of 40 nM mScarlet I-Plk4 (Full length) in the presence of PEG. Scale bar, 2 μm. c HeLa cells expressing GFP-Plk4 (Kinase) and GFP-Plk4 (Kinase + L1). Cells were treated with DMSO or 100 nM Centrinone (20 h). Scale bar, 10 μm; inset, 2 μm. d Fusion event of GFP-Plk4 (Kinase + L1)-induced cytoplasmic condensates. Scale bar, 2 μm. e, f FRAP analysis of cytoplasmic GFP-Plk4 (Kinase + L1) condensates (e) and centriolar GFP-Plk4 (Full length) (f) in HeLa cells (n = 10 and 15 cells, respectively). Cells were treated with DMSO, 100 nM centrinone, 10 μM MG132 or 10 μM MLN4924 for 5–6 h respectively. g Amino acid sequence of mutation sites in Plk4. Red letters, mutation sites; Gray background, degron motif. h, i FRAP analysis of cytoplasmic GFP-Plk4 (Kinase + L1) mutant condensates (h) and centriolar GFP-Plk4 (Full length) mutants (i) in HeLa cells (n = 15 cells for each experiment). For e, f, h, i, intensities were normalized with the average of three pre-bleach signals. Graph shows mean ± SD from two (e) or three (f, h, i) independent experiments. Scale bar, 1 μm. j Schematic of the material properties of Plk4 regulated by autophosphorylation. Source data are provided as a Source Data file

Fig. 3

Condensation properties of Plk4 regulate centriolar copy number. a Amino acid sequence of mutation sites in Plk4. Red letters, mutation sites; Gray background, degron motif. b FRAP analysis of cytoplasmic GFP-Plk4 (Kinase + L1) condensates (b) and centriolar GFP-Plk4 (Full length) (c) in HeLa cells (n = 12 and 13 cells, respectively). For b and c, intensities were normalized with the average of three pre-bleach signals. Graph shows mean ± SD from two (b) or three (c) independent experiments. d Centriole over-duplication assay. Plk4 (Full length)-3xFLAG was exogenously expressed under the CMV mutant promoter in siPlk4-treated HeLa cells. Cells were transfected with siRNA (24 h) and then with plasmid (48 h). Cells were stained with anti-CP110 antibodies as a centriole marker. Scale bar, 10 μm; inset, 2 μm. Percentages of cells which have >4 CP110 foci were calculated. Graph represents mean percentages ± SD (siControl n = 146, siPlk4 + Emp n = 159, siPlk4 + WT n = 153, siPlk4 + 6A n = 153, siPlk4 + 7A n = 166, siPlk4 + 8A1 n = 155, siPlk4 + 8A2 n = 148, siPlk4 + 9A n = 151, siPlk4 + 10A n = 157, siPlk4 + KD n = 151 cells) from three independent experiments. e Summary of the correlation between the condensed Plk4 dynamics and frequencies of centriole overduplication. Source data are provided as a Source Data file

Fig. 4

A single focus of autophosphorylated Plk4 is generated prior to STIL-HsSAS6 loading. a Schematic of Plk4 distribution around mother centrioles in centriole duplication. b 3D-SIM images of centrioles immunostained with the indicated antibodies in HeLa cells. Cep152 was costained and used for a marker of mother centriole wall. Scale bar, 0.3 μm. c Quantification of centriolar Plk4, Plk4pS305, STIL, and HsSAS6 intensities during G1 phase. Cells were synchronized with thymidine and nocodazole. Mitotic cells were collected and released from nocodazole. Intensities were normalized with the average intensity at 9 h after nocodazole washout. Graph shows mean ± SD of centriolar intensities (For each time point (2, 3, 4, 5, 7, and 9 h), Plk4 n = 94, 92, 99, 106, 108, and 102 centrioles, Plk4pS305 n = 94, 92, 99, 106, 108, and 102 centrioles, STIL n = 72, 91, 97, 92, 98, and 103 centrioles, HsSAS6 n = 99, 100, 102, 107, 103, and 102 centrioles) from two independent experiments. See also Supplementary Fig. 9. d 3D-SIM images of centrioles immunostained with antibodies against HsSAS6, Plk4pS305, and Cep152 in HeLa cells. Relative intensities of each protein were quantified along Cep152 ring and maximum peaks of Plk4pS305 intensity were set to 0°. Intensities were normalized to each maximum intensity. Graph shows mean intensities ± SD (n = 11 and 10 centrioles, respectively). Scale bar, 0.3 μm. e, f HeLa cells were transfected with siPlk4 and siHsSAS6 (24 h) and then transfected with plasmid (26 h) expressing Plk4-3xFLAG under the CMV mutant promoter. After plasmid transfection, cells were arrested at G1 phase with Lovastatin (10 μM, 20 h). e Quantification of centriolar Plk4pS305 intensities in cells expressing Plk4-3xFLAG WT or 7A. Black bars indicate average intensities (n = 54 and 53 centrioles, respectively). Intensities were normalized with the average intensity of the WT. Representative data of two experiments. ****p < 0.0001 (Mann–Whitney U test). f Representative images of centrioles immunostained with the indicated antibodies. Images were obtained by TCS SP8 HSR system with deconvolution. Scale bar, 0.3 μm. Source data are provided as a Source Data file

Fig. 5

Autonomous activation of Plk4 that stems from Plk4 self-condensation drives centriolar loading of STIL-HsSAS6. a Scheme of the experimental condition. G1-arrested HeLa cells (10 μM Lovastatin (Lov) (see also Supplementary Fig. 11)) were treated with 100 nM centrinone (Cent) and then washed out (WO) the centrinone. Cells were continuously treated with Lovastatin after centrinone washout. b, c Representative images of immunostained HsSAS6 and Plk4pS305 (b), STIL and Plk4 (c) in the indicated condition. Scale bar, 1 μm. Quantification data show centriolar Plk4pS305, HsSAS6, STIL, and Plk4 intensities in the indicated condition. Gray and pink dots represent the data from siControl and siHsSAS6-treated cells, respectively. Black bars show mean values (for Plk4pS305 and HsSAS6: siControl n = 166, 161, 171, 154, and 160 centrioles, siHsSAS6 n = 102, 105, 118, 100, and 111 centrioles) (for STIL and Plk4: siControl n = 152, 141, 138, 144, and 138 centrioles, siHsSAS6 n = 109, 96, 104, 105, and 102 centrioles) from two independent experiments. Intensities were normalized to the average intensity of the siControl Centrinone (−) condition. d Schematic summary of the correlation between centriolar phospho-Plk4 (pS305) and STIL-HsSAS6 intensity. Source data are provided as a Source Data file

Fig. 6

Hypothetical model: self-organization of Plk4 directs a single daughter centriole formation per mother centriole. Schematic illustration of hypothetical model. Increasing centriolar Plk4 concentration after mitosis drives condensation and autonomous activation of Plk4 prior to STIL-HsSAS6 loading. Self-organization property of Plk4 driven by the regulated condensation induces spatial pattern formation of autophosphorylated Plk4, which results in a bias formation of activated Plk4. The single focus is a possible target site for STIL-HsSAS6 loading. Model 1: active and inactive Plk4 species that have different diffusion coefficients interact with and generate biased distribution of Plk4 around the mother centriole through trans-autophosphorylation. Inactive Plk4 stably self-assembles onto centrioles, while high concentration of Plk4 promotes autonomous activation of Plk4 through trans-autophosphorylation. Activated Plk4 trans-autophosphorylates inactive Plk4 to become active state and suppresses stable self-assembly of inactive Plk4. The interaction mode of these two components is similar to Turing’s reaction-diffusion model or its analog, the lateral inhibition model. Model 2: active Plk4 proteins are phase-separated into liquid-like droplets around mother centrioles and they eventually form a single and large droplet of active Plk4 through coalescence or Ostwald ripening. Surface tension of Plk4 droplet may work as a key force shaping asymmetric structure of Plk4 around mother centrioles. These two models may cooperatively work for spatial pattern formation of Plk4