The semi-conservative centrosome duplication in cycling cells gives rise to a centrosome composed of a mother and a newly formed daughter centriole1. Both centrioles are regarded as equivalent in their ability to form new centrioles and their symmetric duplication is crucial for cell division homeostasis2,3,4. Multiciliated cells do not use the archetypal duplication program and instead form more than a hundred centrioles that are required for the growth of motile cilia and the efficient propelling of physiological fluids5. The majority of these new centrioles are thought to appear de novo, that is, independently from the centrosome, around electron-dense structures called deuterosomes6,7,8. Their origin remains unknown. Using live imaging combined with correlative super-resolution light and electron microscopy, we show that all new centrioles derive from the pre-existing progenitor cell centrosome through multiple rounds of procentriole seeding. Moreover, we establish that only the daughter centrosomal centriole contributes to deuterosome formation, and thus to over ninety per cent of the final centriole population. This unexpected centriolar asymmetry grants new perspectives when studying cilia-related diseases5,9 and pathological centriole amplification observed in cycling cells and associated with microcephaly and cancer2,3,4,10.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
We thank P. Rostaing of the IBENS Electron microscopy facility for making the correlative 3D-SIM and electron microscopy experiments possible. We also thank the Imaging Facility of IBENS, which is supported by grants from Fédération pour la Recherche sur le Cerveau, Région Ile de France DIM NeRF 2009 and 2011 and France-BioImaging. We wish to thank L. Sengmanivong from the Nikon Imaging Centre at Institut Curie-CNRS. We are grateful to L. Goldstein for providing us with the Kif3a mutant mice. We thank C. Janke for the GT335 antibody, E. A. Nigg for the Cep164 and Plk4 antibodies, T. K. Tang for Cep120 and Cpap antibodies, X. Yan and X. Zhu for Deup1 and Cep152 antibodies, J. Azimzadeh for the Poc5 antibody and X. Morin for his gift of the pCAAGS-Cen2–GFP-mCherry plasmid. We also thank M. Bornens, B. Guirao, J. B. Boulé, C. Janke and J.-F. Brunet for their comments on the manuscript and all the members of the Spassky laboratory for discussions. This study was supported by the CNRS, the ENS, INSERM, an FPGG grant, ANR award ANR-12-BSV4-0006 CILIASTEM, ARC award PJA-20131200184, La ligue contre le cancer-comité de Paris RS14/75-88, the ‘Investissements d’Avenir’ program of the French Government and implemented by the ANR (referenced ANR-10-LABX-54 MEMO LIFE and ANR-11-IDEX-0001-02 PSL* Research University), a start-up grant from the City of Paris and a start-up grant from the FRM. A.A.J. and A.-I.L. received fellowships from the French Ministry of Higher Education and Research.
Extended data figures
Apical (upper panel) and side (lower panel) views of a time-lapse sequence showing Cen2-GFP dynamics (63X magnification, ∆t= 40 minutes) during the maturation of an ependymal progenitor in vitro. Note that halos appear within the centrosome region (14:00-30:40) and transform into flower-like structures (31:20-34:00) before the simultaneous release and apical docking of all the centrioles (34:00-72:40). White arrowheads point to the centrosomal centrioles of the ependymal progenitor when discernible. See Fig. 1(d-h). Time is in hh:mm. Scale bar: 2μm.
Apical (upper panel) and side (lower panel) views of a time-lapse sequence showing Cen2-GFP dynamics (63X magnification, ∆t= 40 minutes) during the maturation of an ependymal progenitor in vitro, starting from the halo formation stage (00:00-04:00). Halos transform into flower-like structures (04:40-12:40) prior to the simultaneous release and apical docking of all the centrioles (12:40-32:40). Cd24 staining at 87:20 shows that basal bodies have nucleated cilia. White arrowheads indicate the centrosomal centrioles of the ependymal progenitor when discernible. See Extended Data Fig. 2(a-b). Time in hh:mm. Scale bar: 2μm.
Apical view of a time-lapse sequence showing Cen2-GFP dynamics (100X magnification, ∆t= 30 minutes) during the process of halo formation. Note that 6 halos have already formed and are numbered in the cytoplasm. White arrowheads indicate centrosomal centrioles. Red arrowhead indicates the 7th and last halo formed from a centrosomal centriole (01:00-04:30). After the last halo is released, the 7 halos transform into flowers (numbered at 12:30). Centrioles become visible on the walls of both centrosomal centrioles (yellow arrowheads). See Fig. 2a for single 0.7 μm z-planes of halo formation. Time in hh:mm. Scale bar: 2μm.
Twelve numbered halos have already formed. White arrowheads indicate centrosomal centrioles. Red arrowhead indicates the 13th and last halo formed from a centrosomal centriole (01:00-08:30). All halos then transform simultaneously into flowers (numbered at 08:30). Centrioles become visible on the wall of both centrosomal centrioles (yellow arrowheads). Centrioles detach simultaneously from both centrosome and deuterosome platforms from 10:30. Note that at 04:30, the halo is going under the second centrosomal centriole. Centrin aggregates are occasionally observed in Cen2-GFP cells. See Extended Data Fig. 2c for single z-planes of halo formation. Time in hh:mm. Scale bar: 2μm.
3D-SIM z-planes (0.1μm) spanning a halo and showing the raspberry-like organization of 20 procentrioles (green: Cen2-GFP, red: SAS-6). Dashed circles (left) outline the Cen2+ subunits and dashed rectangles (right) outline the Cen2+/Sas-6+ procentrioles that are oriented parallel to z-plane. z labeling is in μm.
Time-lapse sequence of the Cen2-GFP/KusabiraOrange-Deup1 dynamics (100X magnification, Δt= 30 minutes) during centriole amplification showing that halos and flowers display a Deup1+ core. Ten halos are already cytoplasmic, some of which are superimposed due to z-stacking. White arrowheads indicate centrosomal centrioles. Magenta arrowhead indicates the 11th and last halo formed from a centrosomal centriole (00:00-01:30). After the last halo is formed, all the halos transform simultaneously into flower-like structures by 05:00. Cyan arrowheads indicate procentrioles that become visible on the walls of both centrosomal centrioles. Centrin aggregates are occasionally observed in Cen2-GFP expressing cells. Time in hh:mm. Scale bar: 2μm.
Cen2-GFP time-lapse imaging (100X magnification, Δt= 30 minutes) at the end of the flower stage showing the synchronized release of procentrioles (t=00:00-03:00) from both centrosomal centrioles and deuterosomes (¯t=2.5±0.85h in 9 cells from 5 independent experiments). Released centrioles have migrated to the apical membrane by 20:30. White arrowheads point to the centrosomal centrioles when discernible. Halos appearing at -01:30 in the lower right corner and flowers appearing at 11:00 are from neighboring cells. Time in hh:mm. Scale bar: 2μm.