Trans-generational inheritance of centromere identity requires the CENP-A N-terminal tail in theC. elegansmaternal germ line

Abstract

Centromere protein A (CENP-A) is a histone H3 variant that defines centromeric chromatin and is essential for centromere function. In most eukaryotes CENP-A-containing chromatin is epigenetically maintained, and centromere identity is inherited from one cell cycle to the next. In the germ line of the holocentric nematode Caenorhabditis elegans, this inheritance cycle is disrupted. CENP-A is removed at the mitosis-to-meiosis transition and is established de novo on chromatin during diplotene of meiosis I. Here we show that the N-terminal tail of CENP-A is required for the de novo establishment of centromeres, but dispensable for centromere maintenance during embryogenesis. Worms homozygous for a CENP-A tail deletion maintain a functional centromere during development, but give rise to inviable offspring because they fail to re-establish centromeres in the maternal germ line. We identify the N-terminal tail of CENP-A as a critical domain for the interaction with the conserved kinetochore protein KNL-2, and argue that this interaction plays an important role in setting centromere identity in the germ line. We conclude that centromere establishment and maintenance are functionally distinct in C. elegans.

Mitotic chromosome condensation requires phosphorylation of the centromeric protein KNL-2 inC. elegans

“…C. elegans germ cells move through the gonad from distal to proximal zone as they progress through meiosis. CENP-A and KNL-2 are present on chromatin in the mitotically proliferating zone (distal zone), then they are removed at the onset of meiosis (transition zone) and associate with chromatin again in diplotene (proximal zone) (Gassmann et al, 2012; Maddox et al, 2007; Prosée et al, 2020). We investigated chromosome morphology in the S772A/S784A meiotic germ cells and found that it was altered in the proximal zone of the germ line, but seemed unaffected in the pachytene zone (Fig.…”

“…Considering that condensin II achieves a stable chromatin association during prophase in human cells (Gerlich et al, 2006), we favour the hypothesis of condensin II being less efficiently recruited to chromatin in the S772A/S784A strain. It has recently been observed that when KNL-2 loses its interaction with CENP-A, it fails to localise to chromatin, and chromosome condensation becomes defective (Prosée et al, 2020). This suggests that KNL-2 chromatin association precedes condensin II recruitment.…”

“…elegans germ line. KNL-2 is present in the mitotic zone, then removed at the entry to meiotic prophase and reappears when the cells reach the diplotene stage of meiotic prophase (Prosée et al, 2020). The condensin II complex, although present in all germline nuclei, is chromatin-bound only in the mitotic zone, diplotene and diakinesis (Chan et al, 2004).…”

“…CENP-A deposition is regulated by KNL-2, a M18BP1 homologue, which acts as a loading factor in C. elegans (Maddox et al, 2007), and LIN-53, a RbAp46/48 homologue (Lee et al, 2016). KNL-2 and CENP-A interact directly through CENP-A N-terminal tail and are dependent on one another for chromatin binding (de Groot et al, 2021; Maddox et al, 2007; Prosée et al, 2020). They exhibit a similar localisation pattern throughout the cell cycle and have overlapping genomic distributions (Gassmann et al, 2012; Maddox et al, 2007).…”

The N-terminal tail of C. elegans CENP-A interacts with KNL-2 and is essential for centromeric chromatin assembly

“…Thus, whether the N-tail-KNL-2 interaction reported here continues to be required to load CENP-A HCP-3 after de novo loading in the first embryonic division is an important question. A recent study reported that embryos homozygous mutant for N-tail deleted CENP-A HCP-3 , derived from a heterozygous parent and thus having maternally loaded fulllength CENP-A HCP-3 , developed into adults; by contrast, embryos homozygous null mutant for CENP-A HCP-3 derived from a heterozygous parent did not (Prosée et al, 2020). This observation indicates that the N-tail deleted CENP-A HCP-3 supports development following depletion of maternally loaded full-length CENP-A , suggesting that the N-tail may not be essential for loading later in embryogenesis.…”