Abstract
Centrioles play critical roles in the organization of microtubule-based structures, from the mitotic spindle to cilia and flagella. In order to properly execute their various functions, centrioles are subjected to stringent copy number control. Central to this control mechanism is a precise duplication event that takes place during S phase of the cell cycle and involves the assembly of a single daughter centriole in association with each mother centriole . Recent studies have revealed that posttranslational control of the master regulator Plk4/ZYG-1 kinase and its downstream effector SAS-6 is key to ensuring production of a single daughter centriole. In contrast, relatively little is known about how centriole duplication is regulated at a transcriptional level. Here we show that the transcription factor complex EFL-1-DPL-1 both positively and negatively controls centriole duplication in the Caenorhabditis elegans embryo. Specifically, we find that down regulation of EFL-1-DPL-1 can restore centriole duplication in a zyg-1 hypomorphic mutant and that suppression of the zyg-1 mutant phenotype is accompanied by an increase in SAS-6 protein levels. Further, we find evidence that EFL-1-DPL-1 promotes the transcription of zyg-1 and other centriole duplication genes. Our results provide evidence that in a single tissue type, EFL-1-DPL-1 sets the balance between positive and negative regulators of centriole assembly and thus may be part of a homeostatic mechanism that governs centriole assembly.
Centrioles are cylindrical microtubule-based organelles that direct formation of centrosomes and cilia (Winey and O’Toole 2014). Dividing cells possess one or two pairs of centrioles, with each pair containing a newer (daughter) centriole oriented orthogonally to an older (mother) centriole. In this cellular context, centriole pairs are enveloped by a proteinaceous matrix called the pericentriolar material or PCM, thereby forming centrosomes, the cell’s primary microtubule-organizing center (MTOC). Centrosomes mediate intracellular transport, establish cell polarity, and organize the poles of the mitotic spindle to aid in the segregation of chromosomes. In noncycling cells, centrioles shed their PCM and move to the plasma membrane where the mother centriole serves as a basal body to organize cilia and flagella, structures important for cell motility and cell signaling.
Given the crucial roles of centrioles in both cycling and noncycling cells, it is not surprising that aberrations in centriole number or structure have been linked to disease. An excess number of centrioles is found in many different types of tumor cell, where they can disrupt spindle structure and chromosome segregation leading to aneuploidy (Ganem et al. 2009). Excess centrioles can also interfere with cilia-based cell signaling (Mahjoub and Stearns 2012) and promote cell migration and invasive behavior (Godinho et al. 2015). Thus, excess centrioles can impact the growth of cells in multiple ways. Beyond cancer, defects in centriole structure or number have been linked to several human diseases including autosomal recessive primary microcephaly, primordial dwarfism, and a collection of disorders called ciliopathies (Chavali et al. 2014).
In dividing cells, centriole number is maintained through a precise duplication event in which each mother centriole gives rise to one, and only one, daughter centriole during S phase (Firat-Karalar and Stearns 2014). As each centriole pair will form a spindle pole during the ensuing M phase, stringent control of centriole assembly helps ensure spindle bipolarity and the fidelity of cell division. Forward and reverse genetic studies in the nematode Caenorhabditis elegans have led to the identification of a set of five core factors that are required for centriole duplication (O’Connell et al. 2001; Kirkham et al. 2003; Leidel and Gönczy 2003; Kemp et al. 2004; Pelletier et al. 2004; Delattre et al. 2004; Dammermann et al. 2004; Leidel et al. 2005; Kitagawa et al. 2011a; Song et al. 2011). Functional orthologs of each of these factors have since been identified in other species including flies and humans, thereby establishing the broad evolutionary conservation of the centriole duplication pathway (Leidel et al. 2005; Habedanck et al. 2005; Bettencourt-Dias et al. 2005; Basto et al. 2006; Kleylein-Sohn et al. 2007; Rodrigues-Martins et al. 2007; Vladar and Stearns 2007; Zhu et al. 2008; Kohlmaier et al. 2009; Stevens et al. 2010; Arquint et al. 2012; Vulprecht et al. 2012).
Centriole assembly is initiated by the recruitment of Polo-like kinase 4 (Plk4) to the site of centriole assembly (Dzhindzhev et al. 2010; Cizmecioglu et al. 2010; Hatch et al. 2010; Slevin et al. 2012; Sonnen et al. 2013; Kim et al. 2013; Shimanovskaya et al. 2014). In vertebrates, this step is executed through a direct physical interaction between Plk4 and its centriole receptors SPD-2 and Cep152. A simpler mechanism operates in worms, where SPD-2 is solely involved in recruiting the Plk4 relative ZYG-1(Delattre et al. 2006; Pelletier et al. 2006). ZYG-1/Plk4 then recruits the coiled-coil domain containing proteins SAS-6 and SAS-5/Stil. The molecular details of this step appear species-specific but involve a direct physical interaction between Plk4/ZYG-1 and either SAS-5 or SAS-6, and subsequent phosphorylation (Lettman et al. 2013; Dzhindzhev et al. 2014; Arquint et al. 2015; Kratz et al. 2015; Moyer et al. 2015). At the assembling centriole, SAS-6 dimers oligomerize to form the centriole scaffold, an elegant cartwheel structure in humans and flies or a simpler central tube-like structure in worms (Kitagawa et al. 2011b; van Breugel et al. 2011). Finally, the coiled-coil containing protein SAS-4 is recruited to the nascent centriole and is required for the assembly of the microtubules of the outer wall (Pelletier et al. 2006; Dammermann et al. 2008; Schmidt et al. 2009).
While many of the molecular details of centriole assembly have been elucidated by recent structural and biochemical studies, many mysteries regarding the regulation of this process remain. In particular, it is not clear how a mother gives birth to a single daughter centriole during each round of duplication. Overexpression/overactivation of the core duplication factors ZYG-1/Plk4 or SAS-6 result in the production of multiple daughter centrioles (Habedanck et al. 2005; Peel et al. 2007; Kleylein-Sohn et al. 2007; Basto et al. 2008; Peters et al. 2010), indicating that careful regulation of the levels and/or activity of these factors plays a role in limiting the number of daughters assembled during each round of duplication. More recently, a number of studies have shed light on the importance of posttranslational mechanisms in regulating centriole duplication; both the levels of Plk4/ZYG-1 and SAS-6 are stringently controlled by regulated proteolysis (Strnad et al. 2007; Cunha-Ferreira et al. 2009; Rogers et al. 2009; Puklowski et al. 2011; Peel et al. 2012; Čajánek et al. 2015).
Little is known about how centriole duplication is controlled at the level of transcription. In 1999, Meraldi and colleagues showed that the heterodimeric transcription factor E2F-DP played a role in regulating the reduplication of centrioles in S-phase arrested CHO cells (Meraldi et al. 1999). However, the relevant genes targeted by E2F were not identified. More recently, several isoforms of the E2F transcription factor family (E2F4 and E2F5), along with their binding partner DP and a cell-specific coregulator multicillin, were found to directly activate the transcription of the core centriole duplication factors in multicilliate cells to upregulate centriole biogenesis (Ma et al. 2014). In fact, activation of this transcriptional complex was required for multicilliate cell differentiation. In contrast to the positive role for E2F4 and E2F5 in multicilliate cells, a negative role for E2F3 was demonstrated in mouse embryonic fibroblasts (MEFs). Specifically, inactivation of E2F3, but not other isoforms of E2F, in MEFs resulted in centrosome amplification (Saavedra et al. 2003). These studies show that E2F-DP may play either a positive or negative role in regulating centriole duplication, with the nature of the role appearing to depend upon the cell type and specific isoform of E2F. Here, we show that E2F-DP also plays a role in regulating centriole duplication in C. elegans embryos. Remarkably, we find that E2F-DP plays both a positive and a negative role in a single cell type and propose that E2F-DP1 controls the balance of positive and negative regulators of centriole assembly.
Materials and Methods
Worm maintenance and strains
Worm strains were cultivated using standard practices (Brenner 1974) at 20° on MYOB plates seeded with OP50 Escherichia coli. A complete list of the strains used in this work can be found in Supporting Information, Table S2. Suppression of the zyg-1(it25) phenotype was assayed at 24° or 23.5°, as indicated. Scatter plots displaying suppression data were generated using the Excel templates provided by Weissgerber et al. (2015).
Transgenic worm strains were made using Mos1-mediated single copy insertion (MosSCI) transformation (Frøkjaer-Jensen et al. 2008). pKO113, the zyg-1 transcriptional reporter construct, was generated using Gateway cloning technology (Thermo Fisher Scientific, Inc., Waltham MA) and contained the wild-type zyg-1 promoter (1082 nucleotides upstream of the transcriptional start site), a gfp-his-58 fusion gene, and the zyg-1 3′ UTR cloned into the MosSCI targeting vector pCFJ210 (Frøkjær-Jensen 2012). An identical approach was used to construct pKO114, except that the Quikchange II site-directed mutagenesis kit (Agilent Technologies, Inc., Santa Clara, CA) was used to mutate the three EFL-1-DPL-1 binding sites in the zyg-1 promoter entry clone prior to Gateway cloning.
Mutation identification
Molecular identification of suppressor mutations was accomplished by combining different mapping strategies with whole-genome sequencing. The preparation of genomic DNA, construction of sequencing libraries, and generation of sequence data were essentially as described previously (Wang et al. 2014). Variants were identified using a pipeline of BFAST (Homer et al. 2009), SAMTools (Li et al. 2009), and ANNOVAR (Wang et al. 2010). Mapping plots were generated using R (R Core Development Team, 2015). Candidate suppressor alleles were limited to homozygous (minimum three independent reads, ≥ 85% variant call), nonsynonymous mutations, and filtered to remove variants common to the strain background.
For dpl-1(bs21), the suppressor mutation was mapped by classical three-factor mapping between dpy-10 and unc-4 on linkage group II (Kemp et al. 2007). The strain containing bs21 (OC204) was sequenced, and candidate suppressors in the dpy-10 -unc-4 interval were identified. For mat-3(bs29), the original map position on linkage group II (Kemp et al. 2007) was found to be incorrect (data not shown). The position of this suppressor was reinvestigated, and three-factor mapping with unc-45 and dpy-1 revealed that bs29 was tightly linked to unc-45 on the left arm of chromosome III. The strain containing bs29 (OC184) was sequenced to identify candidate suppressors in the vicinity of unc-45. For efl-1(bs22). A variation of the one-step method for simultaneous mapping and sequencing was employed (Doitsidou et al. 2010). The bs22 suppressor mutation was introgressed into the Hawaiian CB4856 background by backcrossing 10 successive times to a Hawaiian-introgressed zyg-1(it25) strain. Mapping plots of Hawaiian SNPs across the genome revealed (in addition to the interval flanking zyg-1 on chromosome II) gaps on chromosome I (between 2.0–3.0 Mb) and chromosome V (from 15.0 Mb to the right end). The chromosome I interval encompasses a known locus of genetic incompatibility (zeel-1/peel-1 at 2.35 Mb) between the N2 wild-type and Hawaiian strain backgrounds (Seidel et al. 2008) and was not pursued further. Candidate suppressors were identified in the chromosome V interval.
Genome editing
For genome editing, we utilized coCRISPR technology essentially as described (Arribere et al. 2014). Specifically, we designed guide RNAs (gRNAs) using the CRISPR Design tool at http://crispr.mit.edu. gRNA sequences were inserted into the expression plasmid pDD162 (Dickinson 2013) using the Q5 Site Directed Mutagenesis Kit (New England Biolabs, Inc., Ipswich, MA). All oligos used can be found in Table S3. All constructs were sequence verified. For repair templates, we used single stranded oligomers (Paix et al. 2014) synthesized by Integrated DNA Technologies, Inc. (Coralville, IA). For microinjection, we prepared a mixture of two efl-1 gRNA expression plasmids at 50 ng/μl each, purified efl-1 repair template at 30 ng/μl, the dpy-10 coconversion gRNA expression plasmid at 50 ng/μl, and the dpy-10 repair template at 20 ng/μl.
After injection, P0 hermaphrodites were transferred to individual MYOB plates at 20° and allowed to produce an F1 generation. F1 progeny exhibiting a Rol or Dpy phenotype were picked individually to MYOB plates and allowed to lay F3 eggs. F2 adults were then picked individually to a PCR tube, lysed, then screened by PCR for the loss (wt > mut) or gain (mut > wt) of an NlaIV restriction site affected by the bs22 mutation (underlined residues in the efl-1 gRNA sequences in Table S3).
RNAi
RNAi experiments were carried out by feeding worms E. coli-expressing inducible dsRNA constructs as previously described (Kamath et al. 2003). L4 larvae were seeded onto fresh RNAi plates and the effects of RNAi were monitored 12–24 hr from the L4 stage. The L4440 vector (Source Bioscience, Nottingham, UK), expressing dsRNA against the smd-1 gene, was used as a negative control for all RNAi experiments.
Antibodies and quantitative immunoblotting
Embryos were isolated by washing worms off four 10 cm plates and suspending them in a hypochlorite solution (1.65% hypochlorite, 1 M NaOH) for ∼5 min. Once adult worms were dissolved, embryos were rinsed 3 times using M9 buffer, suspended in ∼50 μl 2 x LDS Sample buffer (Life Technologies, Inc., Carlsbad, CA) and heated to 95° for 5 min. Samples were resolved on NuPage Bis-Tris gels (Life Technologies, Inc., Carlsbad, CA) and transferred to nitrocellulose using the i-Blot transfer system (Thermo Fisher Scientific, Waltham, MA). Blots were probed and analyzed using the Odyssey Infrared Imaging System (LI-COR Biosciences, Inc., Lincoln, NE) as previously described (Song et al. 2011).
The following antibodies were used at a dilution of 1:500 – 1:2000: DM1A, an α-tubulin specific antibody (Sigma), α-SPD-2 (Kemp et al. 2004), α-ZYG-1 (Kemp et al. 2007), and α-SAS-4 (Song et al. 2008). The SAS-6 antibody is a polyclonal antibody raised in guinea pigs to a full-length Glutathione-S-Transferase-SAS-6 fusion protein. The antibody was produced by Pocono Rabbit Farm and Laboratory, Inc. (Canadensis, PA) and affinity purified against a full-length Maltose-Binding Protein-SAS-6 fusion protein. Antibody to a SAS-5-derived peptide (N-CPAERERRIREKYARRK-C) was raised in rabbits and affinity purified by YenZym Antibodies LLC, (San Francisco, CA). IRDye secondary antibodies (LI-COR Biosciences) were used at 1:15,000 and membranes were imaged using the Odyssey Infrared Imaging System (LI-COR Biosciences). Bands were normalized to an α-tubulin loading control and quantitated using Image-J software (NIH, Bethesda, MD). Graphs depict the average normalized protein levels from 3 independent experiments.
qRT-PCR
To isolate RNA, L4 worms were transferred to 25° overnight. Approximately 50–100 adult worms were suspended in 200 μl M9 buffer, and RNA was isolated using the Arcturus PicoPure RNA Isolation Kit (Thermo Fisher Scientific, Inc., Waltham, MA) according to the manufacturer’s instructions. RNA was treated with DNAase and then diluted to 20 ng/μl using RNAase-free ddH2O. qRT-PCR reactions were set up in triplicate using 20 ng of template RNA, a QuantiFast SYBR Green RT-PCR Kit from Qiagen (Valencia, CA), and 10 μM primers (see Table S4 for primer sequences). A negative control lacking reverse transcriptase was set up for each RNA template and a nontemplate control was set up for each primer set. Quantitative RT-PCR was performed on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Inc., Hercules, CA). Amplification reactions were carried out using the following program: 10 min at 50°, 5 min at 95° and then 40 cycles (10 sec at 95° and 30 sec at 55°). A melting curve was determined at the end of each PCR run to verify the formation of a single amplicon. Average Ct values were determined using CFX Manager 3.0 Software and the fold change was calculated using the 2ΔΔCt method. For each primer set, the Ct values were normalized to tba-1 RNA levels and then compared to a wild-type calibrator sample. The error bars indicate the SEM of the triplicate set.
Microscopy and live cell imaging
Centriole duplication was monitored using 4D time-lapse microscopy on a Nikon spinning disc confocal microscope as previously described (Peters et al. 2010). Images for comparing the expression of the zyg-1 transcriptional reporters were taken on the same day using the same camera settings.
Data availability
Sequence data are available from the Sequence Read Archive (SRA) under BioProject PRJNA309986.
Results
Mutation of szy-10 suppresses embryonic lethality and restores centriole duplication in the zyg-1(it25) mutant
The szy-10 gene was initially identified as a genetic suppressor of the temperature-sensitive zyg-1(it25) mutant (Kemp et al. 2007). At the nonpermissive temperature of 24°, embryonic centriole duplication fails in the zyg-1(it25) mutant. As a result, the pair of centrioles derived from the sperm separate and establish a bipolar mitotic spindle during the first embryonic division. The absence of centriole duplication during the first cell cycle results in each daughter cell inheriting only a single centriole, which goes on to organize a monopolar spindle in each blastomere of the two-cell embryo (Figure 1A). This failure in centriole duplication also results in a fully penetrant embryonic lethal phenotype at 24°. In contrast to the zyg-1(it25) single mutant, a zyg-1(it25) szy-10 (bs21) double mutant was able to produce a significant number of viable progeny at 24° (Figure 1B). At the slightly higher temperature of 25°, very little suppression was observed in the zyg-1(it25) szy-10(bs21) double mutant (Figure 1B). As the zyg-1(it25) mutant is significantly more impaired at 25°, the failure of suppression at the higher temperature indicates that the szy-10(bs21) mutation does not bypass the requirement for zyg-1 in centriole duplication. Rather, the szy-10(bs21) mutation likely either elevates the residual ZYG-1 activity in the mutant, or alternatively, eases the requirement for ZYG-1 by facilitating execution of the pathway downstream of ZYG-1.