Transcription at the centromere: Opportunity and danger for the maintenance of epigenetic identity

How to make an organism out of a single cell is one of the most amazing processes in biology. Starting with a single fertilized egg, this cell will start to divide many times to give rise to a multicellular animal, plant or fungus. The cell divisions will not stop once fully grown but continue through the lifetime of the organism. This is essential to ensure that body shape and size are maintained, while experiencing natural cell death or caused by interacting with its environment.
Importantly, what is important at the macroscopic level of an organism also plays a role at the cellular level. Developing but also maintaining its identity is an essential task for a cell. Cellular identity is encoded in the genome which comes in the shape of chromosomes. Chromosomes are formed by long fibers of chromatin, made of DNA that is packaged together with proteins called histones. Four different histones H2A, H2B, H3 and H4 form a larger disk-shaped complex called a nucleosome around which the DNA is wrapped giving the chromatin fiber the appearance of beads on a string. During the cell cycle, chromosomes are duplicated in a mother cell and passed on to two daughter cells as cells divide. While most genetic information is encoded in the DNA sequence, it is widely appreciated that chromatin carries information that is passed on independently of the underlying DNA sequence. This form of inheritance is called "epigenetic" and can determine whether certain genes are expressed or define the identity of specific regions of the chromosome. Centromeres are among those specialised regions. They are visible as constrictions in X-shaped chromosomes under the microscope and essential for the separation of chromosomes during cell division.
Centromere identity is determined epigenetically through the presence of a homolog of histone H3, the centromere-specific histone CENP-A (dCENP-A in Drosophila). Centromeric chromatin is composed of interspersed arrays of CENP-A and canonical histone H3 nucleosomes. While H3 is replenished during DNA replication in S-phase, loading of CENP-A in Drosophila and humans takes place in a replication-independent manner from late mitosis to G1. This process requires the removal of so-called placeholder H3-nucleosomes, which have been positioned on centromeric DNA-sequences during the previous S-phase. In recent years, we and other labs have presented evidence that one particular cellular process called transcription could allow for this chromatin remodeling to take place and allow the exchange of H3 by CENP-A.
Cellular processes that require direct DNA contact like DNA replication or transcription induce large-scale chromatin remodeling events to allow the progression of DNA- and RNA- polymerases. The major role of transcription in the genome is to transcribe genes and produce RNA transcripts encoding proteins, transfer or ribosomal RNA. As there are no genes at the centromere, we hypothesise that transcription is instead used to remodel chromatin and destabilise nucleosomes enough to evict H3-placeholders. This would constitute another fascinating example, how nature solves complex problems by repurposing existing toolkits.
Although an intriguing hypothesis, this process is still not understood and raises many questions: How does centromeric transcription distinguish CENP-A nucleosome that should be retained from placeholder H3-nucleosomes that should be removed? How is this process regulated? In a search for binding partners of CENP-A, we previously identified Spt6, a protein involved in recycling histones during transcription. Here, we propose to investigate the molecular details of human and Drosophila Spt6 binding to CENP-A and H3 histones and how transcription at the centromere contributes to CENP-A loading.
By studying this essential evolutionary conserved process, centromeres can be used as a paradigm to understand the underlying mechanism of epigenetic inheritance to preserve the identity of the cell.

The genome is constantly subjected to dynamic changes of actively dividing cells. Cellular processes like transcription induce large-scale chromatin remodeling that involve disassembly of nucleosomes, thereby challenging the stable transmission of epigenetic marks in the genome. Centromeres are among the best studied examples for epigenetic inheritance and are essential for chromosome segregation in mitosis. Centromere identity is propagated by the centromere-specific histone H3-variant CENP-A. While canonical H3-Histones, are loaded during DNA replication, CENP-A is loaded in mitosis and G1. How previously deposited H3-placeholder nucleosomes are removed to provide space for new CENP-A loading remains still unclear. Centromeric transcription in mitosis is evolutionary conserved. Recent evidence from our lab and others suggests that transcription-mediated chromatin-remodeling could evict nucleosomes, yet this also bears the danger of losing previously incorporated CENP-A nucleosomes. Indeed, we could recently show that the transcription elongation factor and histone chaperone Spt6 act to stabilise CENP-A at the centromere. However, it remains unclear how placeholders are evicted, while CENP-A are selectively maintained. We hypothesise that binding affinities between Spt6 and H3-variant/H4 complexes are intrinsically different and phosphoregulated. Using biochemical and biophysical approaches we will characterise Spt6-histone binding in vitro and in vivo to unravel the molecular details of how the chaperone Spt6 and histone phosphorylation are connected in CENP-A and histone variant retention. We will further use established stable Drosophila cell lines to follow wildtype and phospho-residue mutants of old and new H3-histone variants over multiple cells cycles and monitor their effect on centromeric transcription. Finally, to test how transcription contributes to dCENP-A loading over multiple cell generations, we will specifically silence centromeric transcription