Mucociliary clearance and fluid transport along epithelial surfaces are carried out by multiciliated cells (MCCs). Recently, human mutations in Cyclin O (CCNO) were linked to severe airway disease. Here, we show that Ccno expression is restricted to MCCs and the genetic deletion of Ccno in mouse leads to reduced numbers of multiple motile cilia and characteristic phenotypes of MCC dysfunction including severe hydrocephalus and mucociliary clearance deficits. Reduced cilia numbers are caused by compromised generation of centrioles at deuterosomes, which serve as major amplification platform for centrioles in MCCs. Ccno‐deficient MCCs fail to sufficiently generate deuterosomes, and only reduced numbers of fully functional centrioles that undergo maturation to ciliary basal bodies are formed. Collectively, this study implicates CCNO as first known regulator of deuterosome formation and function for the amplification of centrioles in MCCs.
Cyclin O mutations cause congenital mucociliary clearance disorder linked to a reduced number of motile cilia in the airway epithelium. This is caused by malfunctioning deuterosomes that lead to fewer and partially defective centrioles.
Ccno is expressed early in the embryonic node, and in respiratory and ependymal multiciliated cells (MCCs) later on.
MCCs lacking Ccno generate fewer centrioles that often fail to dock to the plasma membrane, resulting in fewer motile cilia.
Deuterosomes of Ccno‐deficient MCCs are enlarged in size, reduced in number and functionally compromised.
These deuterosomes generate fewer and partially non‐functional centrioles.
Cilia are microtubule‐based, hairlike organelles that protrude from the surfaces of cells and are present in the majority of animal phyla (Carvalho‐Santos et al, 2011). While most cells possess one cilium, then referred to as primary cilium, specialized epithelial cells can generate up to several hundred cilia on their surface, then referred to as multiciliated cells (MCCs). Cilia of MCCs exhibit coordinated beating motility that creates a directional flow enabling fluid transport along epithelial surfaces (Fliegauf et al, 2007). Fluid flow directs the distribution of signalling molecules, or generates mechanical forces for fluid transportation, as found in the ventricular system of the brain. Additionally, fluid flow is essential for the clearance of environment‐exposed epithelia by mucus transport, such as in the respiratory tract epithelium.
Human individuals with congenital ciliary motility disorders present with clinical features of recurrent airway infections, infertility, situs ambiguous and increased incidence of hydrocephalus, collectively referred to as primary ciliary dyskinesia (PCD) (reviewed by Knowles et al, 2013). Recently, human cases of congenital mucociliary clearance disorders with predominant respiratory symptoms have been reported that are primarily caused by the reduced generation of multiple motile cilia (RGMC) (Wallmeier et al, 2014). Mutations in the CCNO gene lead to the compromised generation of sufficient cilia numbers on MCCs, while the beating motility of remaining cilia is not generally disrupted (Wallmeier et al, 2014). This study suggested requirements of CCNO function for docking of centrioles to the apical cell membrane. However, a more detailed analysis of CCNO function during the generation of multiple motile cilia may reveal alternative roles of CCNO for cilia biogenesis in MCCs.
During ciliogenesis, centrioles dock to the plasma membrane, thereby forming basal bodies, which serve as platform for the outgrowth of ciliary axonemes (reviewed by Nigg & Raff, 2009). While in monociliated cells the cilium extends from the former mother centriole, MCCs massively amplify centrioles as initiating event of multiciliogenesis (Sorokin, 1968). Two alternative modes of centriole amplification have been described based on ultrastructural studies (Sorokin, 1968). In the canonical pathway of centriole amplification, several procentrioles form in association with the mother centriole, denoted as the mother‐centriole‐dependent pathway (MCD). Additionally, the massive amplification of centrioles is achieved in a mother‐centriole‐independent, acentriolar fashion at so‐called deuterosomes (deuterosome‐dependent pathway, DD). These serve as additional procentriole nucleation centres and are specific for MCCs (Sorokin, 1968). A recent study has demonstrated that deuterosomes are generated in a consecutive manner strictly from daughter centrioles, implicating that predominantly daughter centrioles contribute to centriole amplification in MCCs (Al Jord et al, 2014). Previously, deuterosomes were solely described by their characteristic strongly electron‐dense, ring‐like appearance in transmission electron microscopy and until recently their composition was unknown (Sorokin, 1968). Two studies have elucidated some molecular components of deuterosomes by the identification of DEUP1 and CCDC78 as proteins that specifically localize to deuterosomes, but not to the foci of mother centrioles during procentriole formation (Klos Dehring et al, 2013; Zhao et al, 2013). Forced expression of Deup1 leads to amplification of centrioles, and bacterially expressed, recombinant DEUP1 forms spherical structures reminiscent of deuterosomes in MCCs (Zhao et al, 2013). To date, the knowledge about regulation of deuterosome formation, number and control of their function during centriole amplification remains limited. However, two recent studies have identified the transcription factor MULTICILIN (Mcidas) as direct and broad regulator of the transcriptional programme responsible for DD‐ciliogenesis in MCCs, including the positive regulation of Ccno (Ma et al, 2014). Additionally, it is assumed that centriole formation via deuterosomes is at least partially controlled by the same set of known factors for centriole biogenesis of cycling cells, such as PLK4 (Kleylein‐Sohn et al, 2007), SAS‐6 (Strnad et al, 2007) and CEP152 (Blachon et al, 2008). However, the precise roles of these players have yet to be identified in MCCs (Tang, 2013).
In this study, we describe highly MCC‐specific expression of Ccno from early phases of centriole amplification and ciliogenesis during embryonic development and in differentiating MCCs in culture. Using targeted genetics, we analysed Ccno‐deficient mice, which show characteristics of PCD‐like phenotypes, as a result of significantly reduced cilia numbers of MCCs. Impaired ciliogenesis of Ccno‐deficient MCCs is caused by functional defects of deuterosome‐mediated centriole amplification. Reduced numbers of deuterosomes are formed and newly generated centrioles exhibit defects of maturation and localization, resulting in reduced numbers of properly docked and fully functional ciliary basal bodies. However, Ccno does not seem to impact on the MCD mode of centriole amplification. In summary, we demonstrate that Ccno plays critical roles in the regulation of deuterosome formation and function, required for amplification of functional centrioles and thus the generation of multiple motile cilia in MCCs.
Ccno is specifically expressed in MCCs and the embryonic node
In an expression screen that aimed for identifying transcripts with regional specific expression during cell lineage formation in the embryonic day 7.5 (E7.5) mouse embryo, we discovered highly regional specific transcription of the Ccno gene in cells of the most anterior tip of the primitive streak and in the embryonic node (Fig 1A). The node is the transient organizing structure that establishes the left‐right body axis in a cilia‐dependent manner. Pit cells in the centre of the node carry motile monocilia that generate a leftward fluid flow across the ventral node surface, required for establishing left‐right asymmetry of the body axis (Shiratori & Hamada, 2006). Ccno expression was exclusively found in pit cells, but was excluded from the crown cells that constitute the margins of the node and carry non‐motile cilia (Fig 1A).
For the robust detection of Ccno gene expression, we established a reporter mouse line by blastocyst injection of ES cells generated by the EUCOMM consortium (Skarnes et al, 2011), which carry a LacZ reporter allele in intron 1 of the Ccno gene (Fig 1B and Supplementary Fig S1A). LacZ expression from this CcnoLacZ reporter allele (referred to as CcnoTA hereafter) recapitulates endogenous mRNA expression (Fig 1A and Supplementary Fig S1C) and was used for successive analyses of Ccno expression. At later stages of development, Ccno expression was restricted to epithelial tissues that share as common feature the presence of multiciliated cells (MCCs) (Fig 1C). Ccno‐expressing epithelia include the ependymal epithelium lining the ventricular surfaces of the brain (Fig 1C), the plexus choroideus within the ventricular lumen (Fig 1D), and the epithelium of the upper (snout epithelium; Fig 1E) and lower respiratory tract (trachea and bronchi, Fig 1F). Expression analysis in tracheal epithelium revealed co‐expression of Ccno and Foxj1, a key transcriptional regulator for the generation of motile cilia (Fig 1G) (Chen et al, 1998; Brody et al, 2000). Collectively, these expression patterns suggest specific functions of Ccno in MCCs.
Ccno deletion in mouse causes functional defects of MCCs due to reduced cilia number
To study Ccno gene function, we generated mice with a targeted deletion at the Ccno gene locus by crossing CcnoTA/+ animals to the transgenic Sox2::Cre germline deleter strain (Hayashi et al, 2002) (Supplementary Fig S1A). The deletion of coding exons 2 and 3 resulted in a Ccno null allele configuration (referred to as CcnoRA hereafter). To generate homozygous deletions of the Ccno gene locus, we intercrossed CcnoRA/+ heterozygous animals. CcnoRA/RA mice were born, however were not represented at Mendelian ratio (25%) at weaning age (postnatal day 21; P21). Instead, only 14% of CcnoRA/RA homozygously deleted mutants could be observed (35/246; Fig 2E), while CcnoRA/RA embryos were found at expected frequency of 25% until E17 (Fig 2E), indicating perinatal lethality until P21 in Ccno‐deficient offspring. From the remaining CcnoRA/RA mice, ~57% (20/35) developed a prominent hydrocephalus, leading to abnormal head morphology (Fig 2A and B) and growth retardation. MRI analysis and dissected brains revealed massively enlarged ventricles and a substantial thinning of the cerebral cortex (Fig 2B and C). All CcnoRA/RA animals that developed severe hydrocephalus died within the first 6 weeks of life. Remaining CcnoRA/RA homozygous animals (15/35, 43%) appeared grossly normal and did not suffer from untimely lethality (Fig 2E). Importantly, no obvious signs of laterality defects in Ccno‐deficient embryos or adult animals, and no sneezing or coughing were observed. Male and female animals that were not severely affected by hydrocephalus were fertile at maturity age when used for breeding.
Hydrocephalus in mice is frequently caused by defective ciliary function of ependymal MCCs leading to the reduced intraventricular transport of cerebrospinal fluid (reviewed by Lee, 2013). To investigate additional manifestations associated with disturbed MCC function, we examined paranasal cavities by mucus staining of CcnoRA/RA animals. Obvious mucosal congestion was found in paranasal cavities of all examined CcnoRA/RA animals (n = 3), but not in control littermates from the same cage (n = 3) (Fig 2D and D'). The predominant manifestation of MCC dysfunction in human individuals is insufficient airway clearance leading to recurrent airway infections. Histologicalanalysis of adult Ccno‐deficient lung tissue did not reveal signs of inflammation or mucus congestions of CcnoRA/RA mice that were maintained in a specific pathogen‐free environment (Supplementary Fig S2A and B). However, scanning electron microscopy (SEM) of tracheal epithelium from CcnoRA/RA adult mice showed a pronounced reduction of cilia number on the surface of tracheal MCCs (Fig 2F and F'). In comparison to wild‐types, cilia of Ccno‐deficient MCCs appeared less grouped and extended mostly from the central regions of the apical cell surface, leaving the margins of MCCs undecorated (Fig 2F and F'). Otherwise, single cilia did not show gross morphological abnormalities and the overall cell number of MCCs appeared unchanged (Fig 2F). Immunofluorescence (IF) staining of adult CcnoRA/RA tracheal epithelium for acetylated α‐tubulin, which marks ciliary axonemes, similarly showed reduced staining in accordance with overall reduced number of cilia (Supplementary Fig S2D). The presence of the motor protein dynein outer arm heavy chain 5 (DNAH5) suggested that remaining cilia are motile (Supplementary Fig S2D') (Fliegauf et al, 2005). This assumption was confirmed by video microscopy of tracheal epithelium that revealed ciliary motility of remaining cilia on Ccno‐deficient MCCs (Supplementary Movies S1 and S2).
To uncover ultrastructural details of MCCs, we employed transmission electron microscopy (TEM) of adult tracheal epithelium. Confirming the results from SEM and IF analyses, TEM also revealed considerably reduced numbers of cilia in Ccno‐deficient MCCs (Fig 2G). Instead of properly docked basal bodies, we observed multiple mislocalized and misshaped basal bodies that failed to correctly dock to the apical plasma membrane and to extend ciliary axonemes (Supplementary Fig S3B–E). Additionally, the cytoplasm of Ccno‐deficient cells frequently contained electron‐dense structures that most likely represent remnants of microtubular‐based organelles such as centrioles and/or basal bodies (Fig 2G and Supplementary Fig S3B–E). Ultrastructural analysis of the few cilia that extend from correctly docked basal bodies showed pleiotropic axonemal defects at increased frequencies (Supplementary Fig S3H–J). These included microtubular alterations, such as singlet (Supplementary Fig S3H), or triplet microtubules (Supplementary Fig S3I), abnormal microtubule doublet number and lack of the central tubule pair (Supplementary Fig S3J). However, also axonemes with normal morphology were observed (Supplementary Fig S3G), likely explaining conservation of some ciliary beating motility that could be seen in Ccno‐deficient tracheal epithelium (Supplementary Movies S1 and S2).
Dynamic expression of Ccno suggests roles during early stages of ciliogenesis
To examine if the observed ciliary defects are caused by disturbed ciliary assembly, or result from altered ciliary maintenance or disassembly, we focused our analyses on tracheal and bronchial epithelium during early stages of MCC development. First, we monitored expression onset of Ccno during lung development (Fig 3A–H). Using the CcnoTA reporter allele, we found first expression in tracheal epithelium at E13, coinciding with the early initiation of MCCs (Fig 3E) (Rawlins et al, 2007). During the following developmental stages, Ccno expression extended to more distal bronchi. By E16, expression could be found from the trachea to the terminal bronchioles (Fig 3G and H). Thus, Ccno expression reflects the spatio‐temporal order of MCC development in the respiratory tract (Rawlins et al, 2007; Vladar & Stearns, 2007).
To examine stage‐dependent Ccno expression during ciliogenesis in more detail, we applied mouse tracheal epithelium cell culture (mTEC) techniques (Fig 3I) and compared Ccno mRNA expression with transcription of previously described key components for MCC ciliogenesis in wild‐type mTECs (Zhao et al, 2013). mRNA levels were quantified by qRT–PCR after onset of air–liquid interface culture (ALI) conditions that induces MCC differentiation and ciliogenesis in a semi‐synchronized fashion (Fig 3J). Expression profiles of Deup1, Cep152, Cep63 and Ift57 recapitulated published data sets (Zhao et al, 2013). Ccno mRNA is strongly induced during the first day after onset of ALI conditions and peaked around day 3 (ALI day 3) of differentiation with a 90‐fold induction of mRNA levels compared to day 0, thereby paralleling the expression of Deup1 (Fig 3J). DEUP1 is one of the core components of deuterosomes and becomes upregulated during the early phase of deuterosome formation in mTECs (Zhao et al, 2013). The immediate and strong induction of Ccno expression at early stages of tracheal epithelium and mTEC differentiation thus suggests that Ccno, similar to Deup1, may function during the early steps of centriole amplification.
To test if loss of Ccno function impacts on the previously described transcriptional programme for deuterosome‐dependent (DD) centriole amplification (Ma et al, 2014), we performed qRT–PCRs in control and Ccno‐deficient mTECs (Fig 3K). In wild‐type mTECs, key components of DD centriole amplification are strongly induced during the first 3 days of differentiation and expression declines shortly afterwards. Recently, it was demonstrated that MULTICILIN is the key transcriptional regulator for the expression of multiple factors for DD centriole amplification, including Deup1, Ccno, Cep152 and Cp110 (Ma et al, 2014). In Ccno‐deficient mTECs, Multicilin and all tested downstream targets were strongly induced (Fig 3K) during the first 3 days of differentiation similar to the wild‐type cultures. However, unexpectedly Ccno‐deficient mTECs failed to downregulate Multicilin and other components of the DD pathway after the first induction phase and showed further increasing expression levels until day 7. Interestingly, no increase of Cep63 expression, a key component for mother‐centriole‐dependent (MCD) centriole formation, was observed. This could indicate that loss of Ccno function specifically impacts on deuterosome‐mediated centriole amplification, but not on the MCD pathway.
Ccno‐deficient MCCs show ultrastructural and functional defects of deuterosomes and forming centrioles
To analyse functional consequences of Ccno deletion at early stages of MCC ciliogenesis in more detail, we performed SEM of tracheal tissue at E17. In contrast to SEM analysis of adult trachea (Fig 2F), we observed a complete absence of multiple cilia on the epithelial surface of Ccno‐deficient trachea at E17 (Fig 4A–B'). However, ciliogenesis was not generally ablated at this stage in CcnoRA/RA MCCs, since cells with mono‐ or few cilia with normal morphology could be detected (Fig 4B'). The absence of multiple cilia in E17 tracheal MCCs was also reflected by a significant reduction of acetylated α‐tubulin, marking ciliary axonemes (Fig 4C–E). To investigate the underlying cause for the failure to generate multiple cilia, we performed TEM analysis of E17 tracheal epithelium. As expected from SEM, we observed only very few cilia and similarly only few basal bodies that had properly docked to the apical cell membrane in Ccno‐deficient MCCs (Supplementary Fig S4A and B). At E17, wild‐type MCCs frequently exhibited deuterosomes with forming procentrioles that arrange in a symmetrical fashion surrounding the deuterosome cortex (Fig 4F, F' and H). In striking contrast, deuterosomes of Ccno‐deficient MCCs only rarely displayed clearly discernable procentrioles (Fig 4G, G' and I), and the few emerging procentrioles were significantly shorter in comparison to controls (Fig 4J). Additionally, the usual symmetrical orientation of centrioles was lost, showing an uneven distribution around the deuterosomes (Fig 4H and I, and Supplementary Fig S4E and E'). As most obvious morphological alteration, the great majority of deuterosomes of Ccno‐deficient MCCs showed an irregular shape and were significantly enlarged by 42% (Fig 4K) when compared to deuterosomes of wild‐type cells (Fig 4F–I and Supplementary Fig S4). Enlarged deuterosomes often showed more than one electron‐transparent central region, surrounded by the electron‐dense cortex (Fig 4G and G'). In accordance with previous reports (Sorokin, 1968), our TEM analyses confirmed that the deuterosome‐dependent (DD) mode represents the predominant way of centriole amplification in tracheal MCCs and deuterosomes outnumbered mother centrioles with forming procentrioles by far. However, mother‐centriole‐dependent (MCD) centriole amplification could also be found in MCCs of both wild‐type and Ccno‐deficient tracheae at E17 at much lower frequency (Supplementary Fig S4C–D''). In contrast to deuterosomes, these centrioles appeared structurally unaltered and no striking differences between procentrioles of control and Ccno‐deficient MCCs could be observed (Supplementary Fig S4C–D''). This finding is in accordance with our previous observation by qRT–PCR analysis (Fig 3K) of specific requirements of Ccno for DD, but not MCD generation of centrioles.
Formation of deuterosomes and centrioles depends on Ccno function
To investigate the formation of deuterosome and centrioles in Ccno‐deficient MCCs in more detail, we used mTEC cultures followed by IF analysis to mark deuterosomes by staining for DEUP1, and SAS‐6 as an early marker for centriole formation (Strnad et al, 2007) (Fig 5A). Quantification of IF staining at ALI day 3 revealed an average number of 51 DEUP1‐positive deuterosomes with an average diameter of 348 nm in wild‐type MCCs (Fig 5C). In Ccno‐deficient MCCs, deuterosome numbers were significantly reduced to an average of 21 per cell (~60% reduction), while the average deuterosome diameter was enlarged to 515 nm (~48% increase), thus corroborating our previous observations of enlarged deuterosomes by TEM (Fig 4K). This size increase of deuterosomes was further enhanced at later stages of ciliogenesis at ALI day 7 (714 nm), while deuterosome number remained unchanged (19) in Ccno‐deficient MCCs, but decreased in wild‐type MCCs (44) (Fig 5B and C). SAS‐6 staining is normally found at the circumference of DEUP1‐positive deuterosomes at ALI day 3, indicating the early sites of centriole formation (Zhao et al, 2013). In comparison to controls, SAS‐6 staining was grossly reduced in Ccno‐deficient MCCs (Fig 5A), reflecting the generally reduced generation of centrioles, and confirming our previous observation by TEM (Fig 4G and G'). As a result, at later stages of centriole amplification (ALI day 7), Centrin‐positive centrioles that predominantly arrange close to the apical surface of wild‐type MCCs were dramatically reduced in Ccno‐deficient cells, and the majority of centrioles was found mislocalized at positions basal to, or at the level of deuterosomes (Fig 5B). Thus, Ccno function is required for the proper formation of normal numbers, size and morphology of deuterosomes, and as a consequence, early steps of centriole biogenesis and centriole localization are severely compromised in Ccno‐deficient cells.
To further quantify centriole formation and to analyse maturation of centrioles, we performed double IF staining for Centrin and CEP164, which marks maturing centrioles during their transition to basal bodies (Graser et al, 2007). At ALI day 7, wild‐type cells possess ~230 Centrin‐positive centrioles, while centriole number was significantly reduced to ~80 in Ccno‐deficient MCCs (Fig 6A and B). Additionally, the majority of centrioles in Ccno‐deficient MCCs were not decorated by CEP164‐staining, which is normally colocalizing with Centrin at maturing centrioles (Fig 6A). In reverse, IF staining for CP110, which indicates immature centrioles and is normally released from centrioles during maturation and centriole localization to the apical cell surface (Spektor et al, 2007; Song et al, 2014), showed increased and extended staining in Ccno‐deficient cells. In contrast, CP110 is normally found only in few MCCs of wild‐type mTEC cultures at ALI day 7 (Fig 6C and D). Thus, the significantly reduced staining of CEP164 in combination with retained presence of CP110 demonstrates disturbed maturation of the reduced number of centrioles in Ccno‐deficient MCCs. Defective centriole maturation is also accompanied by the failure of apical localization of centrioles as seen in z‐stack projections (Fig 6A and C).
In summary, Ccno‐deficient MCCs fail to generate sufficient numbers of functional deuterosomes and consequently form reduced numbers of centrioles. Additionally, the fewer centrioles formed in Ccno‐deficient MCCs exhibit maturation defects and fail to localize correctly during their transition to basal bodies. As a result, Ccno‐deficient MCCs show grossly reduced numbers of motile cilia on their apical surface, leading to mucociliary clearance deficits in Ccno‐deficient animals.
This study demonstrates functions of Ccno for the deuterosome‐mediated generation of multiple motile cilia in MCCs. Ccno is very specifically expressed in a dynamic fashion in MCCs, peaking during the earliest phase of deuterosome formation and procentriole generation. The genetic deletion of Ccno in mouse leads to the reduced generation of cilia in MCCs, which is most prominent during initial phases of ciliogenesis at embryonic stages. Our ultrastructural and molecular findings suggest specific functions of Ccno for the proper formation of deuterosomes and for the deuterosome‐dependent generation of procentrioles. Additionally, maturation and apical docking of newly formed centrioles are severely compromised in Ccno‐deficient MCCs. As a consequence, Ccno deficiency causes pleiotropic ultrastructural defects of centrioles, basal bodies and ciliary axonemes and it will be interesting to learn which of these phenotypes are a direct consequence, or secondary effect following loss of Ccno function. Additionally, Ccno deficiency leads to maintained upregulation of the transcriptional programme that controls centriole amplification, including the key transcriptional regulator Multicilin (Ma et al, 2014). Maintained expression of factors required for centriole amplification might contribute to the observed phenotype of increased deuterosome size, when the duration of deuterosome formation is prolonged. Furthermore, the maintained upregulation of the transcriptional programme for centriole amplification suggests the existence of a regulatory feedback loop that potentially senses centriole number to transcriptionally limit the generation of new centrioles, once sufficient numbers are reached.
Collectively, our observations in mouse extend the range of CCNO functions to earlier stages of centriole amplification, further then previous studies of the cellular phenotype of human individuals with CCNO mutations had suggested. Those patients similarly present reduced numbers of cilia of MCCs (Wallmeier et al, 2014). However, clinically consequences seem to considerably differ between human and mouse, since hydrocephalus is only rarely found in human individuals and mice do not show obvious respiratory constraints. Similar to human individuals with CCNO mutations (Wallmeier et al, 2014), and in contrast to previously described PCD patients that entirely fail to form motile cilia (Knowles et al, 2013), Ccno deficiency in mouse does not cause situs inversus, despite highly specific Ccno expression in the embryonic node. Elegant studies in mouse had demonstrated that only very few motile cilia in the node are sufficient to break lateral symmetry by unidirectional nodal flow (Shinohara et al 2012). Since the ability to form multiple motile cilia is not generally abolished in Ccno‐deficient embryos, laterality defects are consequently not observed. Further studies will be required to shed light on Ccno functions in the monociliated cells of the embryonic node.
Cyclins most frequently act by regulating cyclin‐dependent kinases (CDKs). While kinases or other molecular targets that are controlled by Ccno remain to be defined, their identification may provide fundamental insights into regulation of deuterosome‐mediated centriole amplification. It is tempting to speculate that Ccno, which is specifically expressed in MCCs, underwent a functional switch from promoting cell cycle progression to adopting a new role during deuterosome‐dependent amplification of centrioles. Both processes, centriole duplication during S‐phase in cycling cells and centriole amplification at deuterosomes of MCCs, use in parts the same regulatory machinery that might be redirected by Ccno functions in MCCs (Klos Dehring et al, 2013; Zhao et al, 2013).
The precise molecular action of CCNO currently remains unresolved. Thus, it will be interesting to learn if the various defects at multiple stages of centriole biogenesis in Ccno‐deficient MCCs are caused by different, possibly independent CCNO functions, or if all defects arise from disturbances of one common cellular process that is coordinated by CCNO. One example of such an overreaching process could be the spatio‐temporal control of microtubular and/or actin‐based trafficking of deuterosome or centriole components. Disturbances hereof could simultaneously impact on the formation of deuterosomes and on the different steps of ciliogenesis, for example the generation, maturation and localization of centrioles. These different steps of ciliogenesis are normally executed in a highly synchronized manner that involves all centrioles of a cell during MCC ciliogenesis (Al Jord et al, 2014). Our data indicate that at least one step, namely the maturation of centrioles, is not fully synchronized in Ccno‐deficient MCCs, since the presence of different maturation stages of centrioles can be observed in the same cell. Further molecular analyses and dynamic imaging of centriole biogenesis in Ccno‐deficient MCCs will allow unveiling Ccno‐dependent regulation of deuterosome function during centriole amplification.
Materials and Methods
Mouse tracheal epithelial cell (mTEC) culture
mTECs were isolated from wild‐type, CcnoTA/+ or CcnoRA/RA mice at 6 weeks of age or older, as described previously (You et al, 2002; Vladar & Brody, 2013). In brief, tracheal epithelial cells were isolated by overnight pronase digestion at 4°C and separated on the following day from contaminating fibroblasts by a 3‐ to 4‐h incubation on a Primaria cell culture plate in mTEC basal medium (Vladar & Brody, 2013). mTECs were seeded on collagen‐coated apical chambers of Transwell‐Clear filters. Proliferation medium (Vladar & Brody, 2013) was applied to the apical and basal chamber of the transwells and cells cultured at 37°C, 5% CO2. Air–liquid interface (ALI) culture conditions were applied after 6 days of culture by removal of the medium from the apical chamber. Additionally, differentiation was initiated by exchange of proliferation medium with differentiation medium (basal medium + 50 nM retinoic acid + 2% NuSerum).
Immunofluorescence staining of mTEC cultures
mTEC cultures were briefly washed with PBS and fixed in 4% PFA for 15 min at RT, washed with PBS containing 0.1% Triton X (PBX) and blocked for 1 h in PBX containing 4% bovine serum albumin (BSA) and 10% foetal bovine serum (FBS). mTEC cultures were incubated with primary antibody in blocking solution for 1.5 h at RT, using the antibodies listed in Supplementary Table S1. After washing, samples were incubated in PBX for 45 min at RT with conjugated secondary antibody as listed in Supplementary Table S1. mTEC filters were mounted with ProLong Diamond Antifade Mountant (Life technologies P36961) and fluorescent images were captured with a Zeiss inverted laser‐scanning microscope (LSM 510 Meta), equipped with a 1.4×/100× Plan‐Apochromat objective. Image stacks were recorded with a z‐distance of 0.38 μm and subjected to Huygens deconvolution, using Huygens remote manager (hrm) software. Displayed images are snapshots taken from the 3‐D surpass display mode of Imaris 7.2.2 software after Huygens deconvolution. For all images used for figure assembly, the brightness was equally increased from 0 to 20, using Adobe Photoshop CS6.
Analysis and statistics for immunofluorescence of mTEC cultures
After Huygens deconvolution, images were analysed with Imaris 7.7.2 software. For calculations of the mean diameter, 7–10 deuterosomes were measured at the z‐plane of their largest diameter in the 2D‐Slice display mode. Mean diameters of n = 10 cells at ALI day 3 (wild‐type and CcnoRA/RA cells) and n = 12 (wild‐type), and n = 14 (CcnoRA/RA) cells at ALI day 7 were measured. Average deuterosome number was analysed in the 3D‐Surpass display mode, using the spot measurement tool. Estimated spot diameter was based on the mean deuterosome diameter calculated per cell in the previous analysis. Mean deuterosome number per cell was calculated from n = 17 (wild‐type) and n = 19 (CcnoRA/RA) cells at ALI day 3, and n = 12 (wild‐type) or n = 14 (CcnoRA/RA) cells at ALI day 7. For analysis of average centriole number, those Ccno‐deficient cells were chosen that showed clearly discernible centriolar structures, due to increased heterogeneity of Centrin IF staining in Ccno‐deficient cells when compared to wild‐type cells. Measurement of centrioles was performed as described for deuterosome number (n = 11 cells for wild‐type and CcnoRA/RA) at ALI day 7. For statistical analysis, a two‐sided Student's t‐test (P‐value indicated on respective graph) was applied and the standard error of the mean (s.e.m.) was calculated for each data set individually.
Scanning and transmission electron microscopy
For electron microscopy, tissues were fixed in 2.5% glutaraldehyde, 2% PFA and 0.1% picric acid in 0.1 M phosphate buffer pH 7.0 followed by washes in 0.1 M phosphate buffer pH 7.0. Processing for transmission electron microscopy (TEM) included postfixation with an aqueous solution of 1% osmiumtetroxide for 1 h, washes in distilled H2O and processing through an increasing acetone series to 100% acetone, followed by infiltration with epoxy resin (Epon 812, Serva, Heidelberg, Germany) starting in a 1:1 mixture of acetone and resin o/n, followed by pure resin for 4 h. Specimens were transferred to rubber moulds and cured at 60°C for 2 days. Resin blocks were trimmed using the Leica EM Trim 2 (Leica Microsystems, Wetzlar, Germany). Semithin sections (approximately 0.5 μm) and thin sections (50–70 nm) were cut on a ultramicrotome (Ultracut S, Reichert, Wien, Austria) with a diamond knife (Diatome, Biel, Switzerland). Semithin sections were stained with toluidine blue to visualize areas for thin sectioning and ultrastructural inspection. Thin sections were transferred to copper grids and stained with uranyl acetate and lead citrate and were examined on a Zeiss EM902 electron microscope operated at 80 kV (Carl Zeiss, Oberkochen, Germany). Digital images were acquired with a side‐mounted 1x2k FT‐CCD Camera (Proscan, Scheuring, Germany) using iTEM camera control and imaging software (Olympus Soft Imaging Solutions, Münster, Germany).
For scanning electron microscopy (SEM), tissues were dehydrated with a graded series of either ethanol or acetone, followed by immersion in CO2 and subsequent critical point drying in a Emitech K850 critical point dryer (Emitech Ltd. Ashford, UK). Specimens were mounted on SEM stubs with adhesive carbon tape (Plano, Wetzlar, Germany) and sputter‐coated with a gold layer (approximately 15–20 nm thickness) using a Bal‐Tec SCD004 sputter coater (Balzers Union Ltd., Balzers, Liechtenstein). Specimens were viewed on a Zeiss DSM960A SEM equipped with digital image scanning and processing systems and software (Point Electric, Halle, Germany) operated at 10–20 kV and in a field‐emission SEM Zeiss Merlin VP compact operated at 10 kV. Images with a size of 512 × 512 and 1000 × 1000 pixels (DSM960A) and 1024 × 768 pixels (Merlin VP compact) were recorded.
Morphometric measurements and quantification of TEM
The diameter of deuterosomes and length of procentrioles attached to deuterosomes were separately analysed on the same sets of TEM micrographs from E17.5 wild‐type (n = 3) and Ccno‐deficient (n = 5) embryonic trachea specimens using the free line measurement tool in the iTEM imaging software. The annotated values from each picture were exported into Microsoft Excel‐compatible tables using the export function of the iTEM software. For deuterosomes, all clearly identifiable deuterosomes on the pictures obtained from the specimens were measured across for their largest diameter, in total n = 41 wild‐type and n = 38 Ccno‐deficient deuterosomes. For procentriole length, the largest visible length of the procentriole perpendicular to the deuterosome was measured and only procentrioles on deuterosomes bearing two or more clearly visible procentrioles were taken into account. A total of 49 wild‐type (in n = 3 independent samples) and n = 47 Ccno‐deficient (in n = 3 independent samples) procentrioles were measured. For statistical analysis, both data sets (wt and Ccno‐deficient) of deuterosome diameters and centriole length were applied to a two‐sided Student's t‐test (P‐values indicated on respective graph).
All mice were housed in the pathogen‐free barrier facility of the University Medical Centre Freiburg in accordance with the institutional guidelines and approval by the regional board.
Additional Material and Methods including tables of antibodies and primers can be found in the Supplementary Information.
MCF and SJA conceived the study, performed experiments and analysed the data. ANB, GK, KT and SSL assisted and supervised some experiments. TM performed IF analyses and video microscopy of tracheal and ependymal epithelium. MF performed, analysed and interpreted electron microscopy. JD and HO helped with data analysis and during writing of the manuscript. MCF assembled the figures, SJA wrote the manuscript, and SJA and MCF edited the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
Supplementary Movie S1
Supplementary Movie S2
We thank Petra Pennekamp, Niki Tomas Loges und Julia Wallmeier for video microscopy and IF; Ute Schulz, Wolfgang Labs and Gerhard Fulda for technical assistance with SEM/TEM; Wilfried Reichardt for MRI analysis; Sarah Dräger and Hannes Engel for help during in situ expression screening; Barbara Müller and Simone Bräg for technical assistance of lentiviral transduction; Roland Nitschke and the Freiburg Life Imaging Centre (LIC) for supervision of confocal microscopy; Erich Nigg and Xiumin Yan for antibodies; and Alexis Hofherr, Miriam Schmidts, Simone Probst, Michael Köttgen and Gerd Walz for thoughtful discussions and comments on the manuscript. This work was supported by the Emmy Noether Programme of the German Research Foundation (DFG) to SJA (AR732/1‐1), BIOSS Centre of Biological Signalling Studies to SJA and the Excellence Initiative of the German Research Foundation (GSC‐4, Spemann Graduate School).
FundingEmmy Noether Programme of the German Research FoundationAR732/1‐1
- © 2015 The Authors