The Cep57-pericentrin module organizes PCM expansion and centriole engagement

Abstract

Centriole duplication occurs once per cell cycle to ensure robust formation of bipolar spindles and chromosome segregation. Each newly-formed daughter centriole remains connected to its mother centriole until late mitosis. The disengagement of the centriole pair is required for centriole duplication. However, the mechanisms underlying centriole engagement remain poorly understood. Here, we show that Cep57 is required for pericentriolar material (PCM) organization that regulates centriole engagement. Depletion of Cep57 causes PCM disorganization and precocious centriole disengagement during mitosis. The disengaged daughter centrioles acquire ectopic microtubule-organizing-center activity, which results in chromosome mis-segregation. Similar defects are observed in mosaic variegated aneuploidy syndrome patient cells with cep57 mutations. We also find that Cep57 binds to the well-conserved PACT domain of pericentrin. Microcephaly osteodysplastic primordial dwarfism disease pericentrin mutations impair the Cep57-pericentrin interaction and lead to PCM disorganization. Together, our work demonstrates that Cep57 provides a critical interface between the centriole core and PCM.

Fig. 1

Cep57 is an evolutionarily conserved component and forms ring-like structures around the mother centriole wall. a Schematic diagrams of H. sapiens Cep57 (HsCep57) and T. cruzi Cep57 (TcCep57). Coiled-coil domains are shown in gray box. The position of evolutionarily conserved domain (PINC motif) is indicated in pink line. b Alignments of the evolutionarily conserved domain (PINC motif) within H. sapiens, M. musculus, G. domesticus, X. laevis, D. rerio, and T. cruzi. Identical residues determined by Clustal Omega are shown in red. Asterisks indicate the residues identical in all aligned sequences; colons: conserved substitutions; periods: semi-conserved substitutions. c Centriolar distribution of Cep57 at different cell cycle stages. HeLa cells were immunostained with antibodies against Cep57 (red) and GT335 (green) and observed by STED microscopy. Scale bar, 1 μm. d STED images representing top views of Cep57 and Cep192 at mother centrioles. Scale bar, 200 nm. e The graph shows radial profiles from the center of the Cep57 and Cep192 rings. The obtained profile was then fitted with a Gaussian curve and the distance between the center of the ring and the peak of the Gaussian curve was defined as the radius. The diameter of the ring was defined as twice the radius. Values are mean distance ± s.d. (n = 10). f Cep57 becomes gradually enriched at new mother centrioles in interphase. HeLa cells were immunostained with antibodies against Cep57 and centrin-2. Arrowheads indicate new mother centrioles (OM old mother, NM new mother). Scale bar, 1 μm. g The signal intensity of Cep57 at new mother centrioles is proportional to that of PCNT. HeLa cells were immunostained with antibodies against Cep57 and PCNT. Scale bar, 1 μm. h Dot plots represent quantification of the signal intensity of Cep57 and PCNT at new mother centrioles (n = 36). i HeLa cells co-expressing HA empty (control), HA-Cep57, or the indicated HA-Cep57 deletion mutants (1 μg) and Histone 2B-RFP (transfection control, 10 ng) were immunostained with antibodies against HA (green), Cep192 (cyan) and RFP (red). Scale bar, 5 μm in the low-magnified view, 1 μm in the inset

Fig. 2

Cep57 is essential for centriole engagement and PCM organization in mitosis. a HeLa cells were treated with siControl or siCep57 and immunostained with antibodies against centrin-2 (green) and Cep192 (red). Left–right arrows indicate precociously disengaged centrioles. b Schematic model of the phenotype of Cep57 knockdown observed in a. Depletion of Cep57 led to PCM disorganization and precocious centriole disengagement. c Histograms represent frequency of mitotic cells with the indicated phenotypes observed in a. Values are mean percentages ± s.d. from three independent experiments (n = 50 for each experiment). d Time-lapse observation of cells upon Cep57 depletion. HeLa cells expressing GFP-centrin-1 were treated with siControl or siCep57 and observed in the presence of SiR-DNA (200 nM). Green and blue represent GFP-centrin-1 and DNA, respectively. Left–right arrows indicate precociously disengaged centrioles. Time zero corresponds to the beginning of nuclear envelope breakdown (NEBD). e Time-lapse observation of unequal distribution of centrioles (UDC) in Cep57-depleted cells. Precociously disengaged centrioles move around in the cells, which sometimes results in UDC. Arrowheads indicate the precociously disengaged centriole moving toward the opposite spindle pole. f Histograms represent frequency of the cells with the indicated phenotypes observed in e. Values are mean percentages ± s.d. from three independent experiments (siControl n = 116, siCep57 n = 119). g Aberrant acquisition of PCM components of precociously disengaged daughter centrioles in Cep57-depleted cells. HeLa cells were immunostained with the indicated antibodies h Ectopic MTOC activity of precociously disengaged daughter centrioles in Cep57-depleted cells. HeLa cells were immunostained with antibodies against α-Tubulin (green) and CP110 (red). Arrowheads indicate the aberrant MTOC activity in Cep57-depleted mitotic cells. i Schematic illustration of the phenotype of Cep57 knockdown observed in g and h. Precociously disengaged daughter centrioles aberrantly acquire PCM components and MTOC activity. All scale bars, 5 μm in the low-magnified view, 1 μm in the inset. Two-tailed, unpaired Student’s t-test was used in c and f to obtain p value. **p < 0.01

Fig. 3

Depletion of Cep57 or cep57 mutations in MVA disease patients cause chromosome segregation errors and aneuploidy. a Chromosomal segregation errors observed in Cep57-depleted cells. HeLa GFP-centrin-1 cells were treated with siControl or siCep57 in the presence of SiR-DNA (200 nM). b Histograms represent frequency of the mitotic cells with the indicated phenotypes in a (siControl n = 116, siCep57 n = 119 from three independent experiments). c Box-and-whiskers plots show the duration from NEBD to AO in a. The ends of the box are the upper and lower quartiles. The median is marked by a vertical line inside the box (two independent experiments, siControl n = 86, siCep57 n = 79). d Pedigree of an MVA disease family. MVA patients 1 and 2 were affected with compound heterozygous cep57 mutations, c.520_521delGA (p.E174fs) and c.915_925dup11 (p.H317fs). Their unaffected father and mother were the carriers of c.520_521delGA and c. 915_925dup11, respectively. e MVA patients’ cells exhibited precocious centriole disengagement and PCM disorganization. MVA patients’ lymphoblastoid cell lines (LCLs) and their mother’s LCL (unaffected control) were immunostained with antibodies against centrin-2 (green) and Cep192 (red). Left–right arrows indicate precociously disengaged centrioles. f Histograms represent frequency of the control and MVA patients’ cells with the indicated phenotypes in e. Values are mean percentages ± s.d. from three independent experiments (n = 30 for each experiment). g Abnormalities in centriole/centrosome number in MVA patients’ cells. MVA patients’ LCLs and their mother’s LCL were immunostained as in e. h Histograms represent frequency of the control and MVA patients’ cells with the indicated number of Cep192 foci in g. Values are mean percentages ± s.d. from three independent experiments (n = 100 for each experiment). i A model for the MVA disease with cep57 mutations. Depletion of Cep57 or cep57 mutations in MVA patients cause PCM disorganization and precocious centriole disengagement and thereby result in chromosomal segregation errors and aneuploidy. All scale bars, 5 μm in the low-magnified view, 1 μm in the inset. Mann–Whitney test was used in c to obtain p value. Dunnett’s multiple comparisons test was used in f and h to obtain p value. **p < 0.01; NS not significantly different (p > 0.05)

Fig. 4

Cep57 is an anchor of the PACT domain of PCNT. a Schematic of Cep57 and the deletion mutants used for co-IP assays. Coiled-coil domains are shown in gray box; the PINC motif in pink line. The region of Cep57 for PCNT-binding is represented in red. The right column shows a summary of the co-IP results. b HEK293T cells co-expressing FLAG empty (control) or FLAG-PCNT and HA-Cep57 were immunoprecipitated (IPed) with FLAG antibodies. c HEK293T cells co-expressing FLAG-PCNT and HA-Cep57 or the indicated deletion mutants were IPed with FLAG antibodies. d Schematic of PCNT and the deletion mutants used for co-IP assays. Coiled-coil domains are shown in gray box, and the PACT domain in blue box. The region of PCNT for Cep57-binding is represented in red. The right column shows a summary of the co-IP results. We noticed that the N-terminal half of PCNT (a.a. 1–1962), which does not contain the PACT domain, also interacted with Cep57. e HEK293T cells co-expressing HA-Cep57 and FLAG-PCNT or the indicated deletion mutants were IPed with FLAG antibodies. f MBP pull-down assay showing the interaction between Cep57 and PCNT fragments in vitro. These bacterially purified recombinant proteins contain the interaction regions that were identified by the co-IP experiments in c and e. Inputs (lane 1 for Coomassie blue staining, 13.3%; lane 1 for western blotting, 1/30,000 volume of lane 1 in the Coomassie blue staining) and affinity-purified protein complexes (lanes 2–5) were subjected to SDS-PAGE, stained (Coomassie blue staining), and analyzed by western blotting. g HeLa cells co-expressing HA empty (control) or HA-Cep57 and GFP or GFP-PCNT 3113–3336 were immunostained with antibodies against HA (red), GFP (green). h HeLa cells were treated with control siRNA or siCep57, followed by transfection with the GFP-PCNT 3132–3226. The cells were immunostained with antibodies against GFP (green), Cep57 (red), and Cep192 (cyan). i Histograms represent quantification of the signal intensity of GFP and Cep57 at old mother centrioles in h (siControl n = 41, siCep57 n = 39). All scale bars, 5 μm. **p < 0.01

Fig. 5

The binding of PCNT to Cep57 at centrioles is required for the proper localization pattern of PCNT to organize mitotic PCM. a Schematic of PCNT and the ΔPACT mutant or MOPD2 disease-related mutants used for co-IP assays. Coiled-coil domains are shown in gray box and the PACT domain in blue box. The region of PCNT for Cep57-binding is represented in red. The right columns show a summary of the co-IP results and rescue experiments, respectively. b HEK293T cells co-expressing HA-Cep57 and FLAG-PCNT full-length or the indicated mutants were IPed with FLAG antibodies. c The signal intensity of HA-Cep57 bands in b was quantified. Values are mean ± s.d. from three independent experiments. Tukey’s multiple comparisons test was used to obtain p value. *p < 0.05; **p < 0.01; NS not significantly different (p > 0.05). d HeLa cells were treated with siControl or siPCNT, followed by transfection with FLAG empty (control), RNAi-resistant (RNAi-R) PCNT or the indicated mutants. The cells were immunostained with antibodies against FLAG (green), Cep192 (cyan) and CP110 (red). e Histograms represent frequency of mitotic cells with the indicated phenotypes observed in d in each condition. Values are mean percentages ± s.d. from three independent experiments (n = 50 for each experiment). f Schematic of PCNT and the PCNT-Cep57 chimera mutant used for PCNT knockdown and rescue experiments. g HeLa cells were treated with siControl or siPCNT, followed by transfection with FLAG empty (control), RNAi-resistant PCNT, or the PCNT-Cep57 chimera mutant. The cells were immunostained with antibodies against FLAG (green), Cep192 (cyan), and CP110 (red). h Histograms represent frequency of mitotic cells with the indicated phenotypes observed in g in each condition. Values are mean percentages ± s.d. from three independent experiments (n = 50 for each experiment). All scale bars, 5 μm in the low-magnified view, 1 μm in the inset. Left–right arrows indicate precociously disengaged centrioles. Tukey’s multiple comparisons test was used in e and h to obtain p value. *p < 0.05; **p < 0.01; NS not significantly different (p > 0.05)

Fig. 6

A speculative model for the role of Cep57 in centriole engagement and PCM integrity during mitosis. Cep57 localizes at the proximal ends of mother centrioles and directly binds to the PACT domain of PCNT. The binding of PCNT to Cep57 at centrioles is required for the proper localization pattern of PCNT to organize the highly ordered PCM structure and to maintain the centriole engagement during mitosis. In the absence of Cep57, the PCNT height that binds along the side of centriole wall is affected, which causes PCNT dispersion and PCM disorganization. PCM disorganization further leads to precocious centriole disengagement during mitosis. Precociously disengaged daughter centrioles aberrantly recruit PCM components and acquire ectopic MTOC activity, which thereby causes chromosome segregation errors and aneuploidy