Fig 1.
Dynamics of KNL-2 and CENP-A in the C. elegans germ line.
Top, cartoon images of an adult hermaphrodite C. elegans and the different stages of the germ line. Bottom, IF images of CENP-A and KNL-2 in adult hermaphrodite germ lines and embryos, showing the removal at the mitosis-to-meiosis transition and the reappearance at the diplotene stage. The white lines indicate where images of the same gonad have been merged. Scale bars represent 20 μm. CENP-A, centromere protein A; IF, immunofluorescence.
Fig 2.
The CENP-A N-terminal tail interacts with the KNL-2 central domain.
(A) Yeast two-hybrid analysis of the interaction between CENP-A and KNL-2. Cartoons of the different CENP-A and KNL-2 fragments used (top), showing the annotated SANTA, SANT/Myb, and CC domains in KNL-2 and HFD in CENP-A. The interaction between the constructs were determined by assaying growth of yeast cells on selective medium (SD-LTH, containing 5 mM 3AT). Vector-only constructs and nonselective plates (SD-LT) were used as negative and positive controls for growth, respectively. Left, full-length KNL-2 interaction with different parts of CENP-A. Center, interaction of the CENP-A N-terminal tail with different regions of KNL-2. Right, interaction of different CENP-A N-terminal tail fragments with the KNL-2 central region. (B) Co-IP experiments with KNL-2 and CENP-A FL or CENP-Aaa100-287 (Δ-tail). In cartoons, CENP-A is shown in yellow and HA in gray. Representative western blot showing IPs of HA-tagged CENP-A FL, or Δ-tail from embryonic extracts. HA-tagged constructs were IPed and detected using an anti-HA antibody. OLLAS-tagged KNL-2 was detected using an anti-OLLAS antibody in input and IP samples. The bar graph shows a quantification of relative KNL-2 enrichment, which was determined as the IP ratio of KNL-2/CENP-A divided by the input ratio of KNL-2/CENP-A, with the interaction for CENP-A FL set to 1 in each experiment. N = 3 independent IP/western blotting experiments. The data underlying all the graphs can be found in S1 Data. CC, coiled-coil; CENP-A, centromere protein A; Co-IP, co-immunoprecipitation; FL, full-length; HFD, histone fold domain.
Fig 3.
The presence of the CENP-A N-terminal tail is dispensable for mitosis during development but required in the germ line.
(A, B) Phenotypic consequences of the CENP-A N-terminal tail deletion in presence or absence of CENP-A FL in the maternal germ line. (A) Quantification of viable offspring of worms carrying CENP-A FL (gray), CENP-A KO mutation (dark blue), or CENP-A tail truncation (Δ-tail, orange) alleles. A balancer allele (hT2) was used to maintain heterozygous strains. In these balancer strains, 10/16 of the offspring are aneuploid, and CENP-A FL homozygotes (1/16) are embryonic lethal due to the balancer alleles and are therefore not shown. Top, cartoon of maternal heterozygous (striped) or homozygous (full) genotypes as well as genotypes of the offspring. Bottom, quantification of the viable offspring as a percentage of the expected hatchlings in the context of the balancer allele. Error bars show the standard deviation of the mean. Connecting lines highlight the most relevant comparisons between the different genotypes. N = 8 to 10 independent broods. (B) Cartoon of an adult germ line, with imaged regions highlighted with black boxes. IF images of both the CENP-A FL and CENP-A Δ-tail proteins in the maternal germ line (diakinesis) and embryos. Left, maternal CENP-A FL/Δ-tail P0 giving rise to either CENP-A FL/Δ-tail or Δ-tail/Δ-tail F1 offspring (indistinguishable as early embryos). Right, maternal CENP-A Δ-tail/Δ-tail F1 giving rise to CENP-A Δ-tail/Δ-tail F2 offspring. Cartoon images with genotype color codes as in (A). (C) Depletion of the CENP-A FL protein during development in CENP-A FL and Δ-tail (F1) homozygous worms, using the AID system. Embryos were placed onto auxin-containing plates for depletion during development. Top, cartoon of the genotypes with color codes as in (A). Bottom, quantification of hatchlings with normal development upon CENP-A FL protein depletion. N = 9 independent broods. (D) Depletion of the CENP-A FL protein in the germ line of CENP-A FL/Δ-tail heterozygous worms (P0), using the AID system. Adults were placed on auxin plates for 4 h. Left, cartoon of maternal genotype and genotypes of the offspring, with color codes as in (A), and quantification of the viable offspring as a percentage of the expected hatchlings in the context of the balanced allele. N = 4–5 independent broods. Right, IF images of both the CENP-A FL and CENP-A Δ-tail proteins in the maternal germ line (diakinesis) and embryos. Adult CENP-A FL/Δ-tail P0 upon AID-mediated depletion of CENP-A FL protein in the germ line, giving rise to CENP-A FL/Δ-tail offspring. Scale bars in (B, D) represent 5 μm. The data underlying all the graphs can be found in S1 Data. AID, auxin-inducible degron; FL, full-length; IF, immunofluorescence; KO, knockout.
Fig 4.
The CENP-A N-terminal tail is required for establishing centromere identity in the germ line.
Cartoons showing the different CENP-A genotypes (FL, gray; KO, dark blue; Δ-tail, orange) as heterozygotes (striped) or homozygotes (full). (A) Maintenance of CENP-A Δ-tail homozygous males by crossing. Percentage CENP-A Δ-tail hetero- and homozygotes among viable male cross-progeny after crossing CENP-A Δ-tail homozygous males with heterozygous CENP-A FL/Δ-tail hermaphrodites. N = 16 independent crosses. (B) Rescue of embryonic lethality of CENP-A KO embryos by zygotic expression of CENP-A FL (left) or Δ-tail (right), introduced by genetic crosses of CENP-A FL or Δ-tail homozygous males to heterozygous CENP-A KO hermaphrodites. Percentages of viable progeny for each genotype are shown. N = 3 independent crosses each. (C) Non-rescue of embryonic lethality of CENP-A Δ-tail F2 embryos by zygotic expression of CENP-A FL, introduced by genetic crossing of wild-type males to homozygous CENP-A Δ-tail F1 hermaphrodites. Percentage of viable progeny is shown. N = 9 independent crosses. (D) Rescue of the embryonic lethality of CENP-A Δ-tail F2 embryos by pulse of CENP-A FL protein expression in F1 adults. (D) Percentage of viable F2 embryos without heat shock and at different time points 0–8 h after heat shock–induced expression of CENP-A FL protein in adult F1 (top). Images of DAPI-stained 1- or 2-cell F2 embryos during anaphase at the indicated time points after heat shock (bottom). N = 3 independent heat shock experiments. Error bars show the standard deviation of the mean. The data underlying all the graphs can be found in S1 Data. CENP-A, centromere protein A; FL, full-length; KO, knockout.
Fig 5.
Centromere identity and function are maintained for one generation upon deletion of the CENP-A N-terminal tail.
(A) Centromere biorientation in metaphase cells. IF images of KNL-2 and CENP-A Δ-tail in metaphase cells in the mitotic zone of the adult hermaphrodite germ line or in early embryos for the different generations (P0, F1, F2) of the CENP-A Δ-tail strain. (B) Still images for the indicated stages of mitosis of recordings in embryos derived from CENP-A FL/FL, FL/Δ-tail, or Δ-tail/Δ-tail maternal germ lines. KNL-2 (GFP), HCP-4 (mCherry), or ROD-1 (GFP) were fluorescently labeled. (C) Still images of recordings in 1-cell embryos derived from CENP-A FL/Δ-tail or Δ-tail/Δ-tail maternal germ lines, showing GFP-histone H2B (chromosomes) and GFP–γ-tubulin (spindle poles). Time is given in seconds relative to NEBD. The arrows indicate the location of the spindle poles. (D) Quantification of spindle pole separation kinetics during the first embryonic cell division in embryos derived from CENP-A FL/Δ-tail or Δ-tail/Δ-tail maternal germ lines. Error bars show the standard deviation of the mean. Scale bars represent 1 μm (A), 4 μm (B), or 10 μm (C). CENP-A Δ-tail homozygote F1 embryos are indistinguishable from heterozygous embryos; therefore, the F1 genotypes are marked with an asterisk in all panels. The data underlying all the graphs can be found in S1 Data. CENP-A, centromere protein A; FL, full-length; IF, immunofluorescence; NEBD, nuclear envelope breakdown; WT, wild-type.
Fig 6.
CENP-A Δ-tail is maintained on chromatin by LIN-53 in absence of the interaction with KNL-2.
(A) Germline-specific depletion of CENP-A FL and KNL-2 using the AID system. IF images of CENP-A FL and KNL-2 in diakinesis nuclei of adult hermaphrodite germ lines, with and without auxin-induced degradation of either KNL-2 (left) or CENP-A FL (right). Bar plots show quantifications of total nuclear fluorescence in diakinesis nuclei. N = 10–18 nuclei total, from 3 independent IF experiments. (B) Chromatin association of CENP-A Δ-tail in presence or absence of KNL-2. IF images and fluorescence quantifications of CENP-A Δ-tail and KNL-2 in diakinesis nuclei of CENP-A Δ-tail F1 homozygous adult hermaphrodite germ lines. KNL-2 was depleted using germline-specific AID. N = 7–8 nuclei total, from one IF experiment. Germline-specific depletion of KNL-2 and CENP-A in (A) and (B) was achieved using a TIR1 with germline- and early embryo-restricted expression. (C) Chromatin association of CENP-A FL and Δ-tail in presence or absence of LIN-53. IF images showing CENP-A FL and CENP-A Δ-tail in diakinesis nuclei and at metaphase in early embryos, for CENP-A FL homozygotes (left) and CENP-A FL/Δ-tail heterozygote P0 adults and CENP-A Δ-tail homozygote F1 embryos (right) (indistinguishable from heterozygous embryos; therefore, the F1 genotype is marked with an asterisk). LIN-53 was depleted by RNAi. Bar plots show quantifications of CENP-A levels in control and LIN-53-depleted animals. N = 26–33 oocytes and 12–26 embryonic metaphase plates from 3 independent IF experiments. In all quantifications, the mean level of the control (no auxin or no RNAi) in each experiment was set to 1. Error bars show the standard deviation of the mean, and the asterisks denote statistical significance, determined by using a Student t test. Scale bars represent 5 μm (diakinesis oocytes) or 1 μm (metaphases). The data underlying all the graphs can be found in S1 Data. AID, auxin-inducible degron; CENP-A, centromere protein A; FL, full-length; IF, immunofluorescence; RNAi, RNA interference; WT, wild-type.
Fig 7.
Model for the establishment and maintenance of centromere identity and the role of the CENP-A N-terminal tail.
(A) Model of centromere establishment and maintenance across generations in C. elegans. Centromeres are established in the proximal germ line and maintained until their removal in the distal germ line of the next generation. The CENP-A N-terminal tail interacts with KNL-2 and is required for establishing centromeres. The presence of CENP-A on chromatin also depends on LIN-53. In the heterozygous P0 worm, the CENP-A FL present in the germ line is sufficient for correct centromere establishment. Upon loss of the CENP-A N-terminal tail, centromeres can be maintained for one generation, a process that is likely mediated by LIN-53. However, in absence of the CENP-A N-terminal tail, centromere establishment fails in the adult F1 generation, leading to mislocalization of CENP-A Δ-tail and KNL-2, and a knl phenotype in the F2 embryos. (B) Two models explaining a mel phenotype, where the function of an essential gene is maintained for one generation despite loss or truncation of the gene. Top, sufficient protein or mRNA is deposited in the embryo from the maternal germ line for it to persist and function during development. Bottom, the protein sets a (chromatin) state in the parental germ line that serves as a memory that persists during development by epigenetic inheritance. The centromere inheritance described in this study follows the epigenetic inheritance model, as indicated by the dotted line. Protein presence is labeled in blue, and epigenetic inheritance in gray. CENP-A, centromere protein A; FL, full-length; knl, kinetochore-null; mel, maternal effect lethal; mRNA, messenger RNA; MTs, microtubules.