Human centromere repositioning activates transcription and opens chromatin fibre structure

Abstract

Human centromeres appear as constrictions on mitotic chromosomes and form a platform for kinetochore assembly in mitosis. Biophysical experiments led to a suggestion that repetitive DNA at centromeric regions form a compact scaffold necessary for function, but this was revised when neocentromeres were discovered on non-repetitive DNA. To test whether centromeres have a special chromatin structure we have analysed the architecture of a neocentromere. Centromere repositioning is accompanied by RNA polymerase II recruitment and active transcription to form a decompacted, negatively supercoiled domain enriched in ‘open’ chromatin fibres. In contrast, centromerisation causes a spreading of repressive epigenetic marks to surrounding regions, delimited by H3K27me3 polycomb boundaries and divergent genes. This flanking domain is transcriptionally silent and partially remodelled to form ‘compact’ chromatin, similar to satellite-containing DNA sequences, and exhibits genomic instability. We suggest transcription disrupts chromatin to provide a foundation for kinetochore formation whilst compact pericentromeric heterochromatin generates mechanical rigidity.

Transcription of a centromere-enriched retroelement and local retention of its RNA are significant features of the CENP-A chromatin landscape

“…However, we cannot distinguish between transcriptional activation of lacO being caused by CAL1 tethering, given that CAL1 is known to interact with RNAPII and FACT (Chen et al, 2015), or being linked to active CENP-A deposition. The latter possibility would be consistent with recent studies in human neocentromeres showing that neocentromere formation is associated with transcriptional activation and increased chromatin accessibility (Murillo-Pineda et al, 2021; Naughton et al, 2022).…”

“…Our finding that de novo centromeres are coupled with transcriptional activation of the underlying DNA specifically in metaphase reinforces the model that CENP-A deposition and transcription go hand in hand. However, we cannot distinguish whether transcriptional activation of lacO is caused by CAL1 tethering, given that CAL1 is known to interact with RNA polymerase II and FACT (Chen et al, 2015), or whether transcription is linked to active CENP-A deposition, consistent with recent studies in human neocentromeres showing that neocentromere formation is associated with transcriptional activation and increased chromatin accessibility (Murillo-Pineda et al, 2021; Naughton et al, 2022).…”

“…Transcripts emanating from centromeres have been observed in a myriad of systems, including budding yeast (Hedouin et al, 2022; Ohkuni and Kitagawa, 2011), human cells (Bury et al, 2020; Hoyt et al, 2022; McNulty et al, 2017; Quenet and Dalal, 2014), frog egg extracts (Blower, 2016; Grenfell et al, 2016), maize (Topp et al, 2004) and marsupials (Carone et al, 2013). Transcription at centromeres has been shown to be coupled to de novo centromere formation (Chen et al, 2015) and neocentromere formation in humans (Chueh et al, 2009; Murillo-Pineda et al, 2021; Naughton et al, 2022). In addition, centromeric transcription is critical for programmed histone exchange in S. pombe (Wu et al, 2022), for the stabilization of newly formed CENP-A nucleosomes in Drosophila cells (Bobkov et al, 2018), and for HAC formation (Cardinale et al, 2009).…”

The role of centromeric repeats and transcripts in kinetochore assembly and function

“…We suggest that KNL2 and CENP‐C target CENH3 nucleosomes via the CENPC motif, and subsequently bind to their DNA. Repositioning of human centromeres is accompanied by RNA polymerase II recruitment, which activates neocentromere transcription (Murillo‐Pineda et al., 2021; Naughton et al., 2022). It can be assumed that all kinetochore components capable of binding to centromeric transcripts in a sequence‐independent manner will also bind non (i.e., neo)‐centromeric transcripts.…”

“…The deletion of the original centromeres can result in the formation of neocentromeres at an alternative position on either chromosome arm via ectopic deposition of a small amount of CENH3. Centromere repositioning can lead to the recruitment of RNA polymerase II to these sites and activation of their transcription (Murillo‐Pineda et al., 2021; Naughton et al., 2022). KNL2 and CENP‐C, as well as other kinetochore proteins capable of binding DNA and RNA in a sequence‐independent manner, can be expected to bind neocentromeres and their transcripts, and this may play a role in the assembly and stability of kinetochores at neocentromeres.…”