Conference Scene: Chromatin, replication and chromosomal stability

Abstract

The Chromatin, Replication and Chromosomal Stability Conference took place on June 20–21 in Stockholm, Sweden. In this article, I outline the broad scientific program of the meeting which reflected the wide diversity in epigenetics research. Distinct histone modifications are linked with specific chromatin structures and intranuclear positioning, thereby impacting replication timing and replication initiation, which in turn are related to gene expression and cell differentiation. Interference in any of these interconnected mechanisms can result in DNA breakage and lead to the activation of repair pathways. The DNA repair mechanisms again are influenced by the chromatin structure. In summary, the conference highlighted the functional implication of epigenetics in chromatin compaction, transcription regulation, replication control and DNA repair. The tight control of all these mechanisms defines the final cellular character.

The symposium was opened with a beautiful presentation by Helen M Blau (Stanford University, CA, USA) highlighting the impact of the cellular environment on nuclear programming. She showed that the differentiated state of a cell is dictated by the balance between different transcription factors and various epigenetic regulators. Both balances require a continuous regulation. How a cell‘s environment impacts on gene expression and epigenetic changes was described by two experimental approaches. First, when human fibroblasts grow in an excess of mouse embryonic stem cells (ESCs), spontaneous cell fusions drive the fibroblast genome into the pluripotent state, including Oct4 gene demethylation and transcription [1]. Second, the elasticity of the cell growth medium impacts on gene expression and cell fate [2]. Therefore, the microenvironment plays an important role in epigenome construction and maintenance.

One major concern raised at the symposium was the regulation of replication initiation and replication timing. David Gilbert (Florida State University, FL, USA) and his coworkers generated genome-wide replication timing profiles of many human and mouse cell lines, ESC differentiation intermediates and cells from diseased patients [101]. In mouse embryonic development, specifically, the early-to-late epiblast transition is accompanied with many early-to-late replication changes, which are accompanied by large-scale spatial re-organizations. As a result genes within these domains become more difficult to reprogram back to the pluripotent state. Replication timing profiles were found to correlate with transcription and active transcription histone marks; however, the highest correlations were obtained with genome-wide maps of long-range chromatin interactions (defined by Hi-C approaches). Gilbert concluded that replication timing changes during development are not necessarily linked to changes in transcription, but rather to changes in chromatin condensation and nuclear positioning [3].

Itamar Simon (Hebrew University-Hadassah Medical School, Jerusalem, Israel) presented an alternative method for genome-wide replication timing profiling in humans and mice, based on copy number changes between G1- and S-phase cells [4]. Like Gilbert, he described a correlation between early replication and histone marks of active transcription. Moreover, he highlighted that a high-GC content is linked to early replication. He explained this observation based on two known phenomena: mutations that occur during late S-phase replication have a preference of GC to AT transitions, and deoxyribonucleoside triphosphate pools are GC-rich in early, and AT-rich in late S-phase. Simon concluded that conserved replication timing patterns may influence regional GC content by affecting the types of mutations that accumulate upon evolution.

Marcel Méchali (Institute of Human Genetics, Montpellier, France) presented genome-wide replication origin mappings based on RNA 5´-DNA primer isolation from mouse and Drosophila melanogaster cells [5]. He concluded that origins preferentially map to promoter and coding regions of active genes, often 600 bp distant from CpG islands. A combination of in silico simulations and combing experiments revealed that nearby origins group into functionally connected replicons, of which only one origin is active in a single cell cycle. Moreover, he observed that embryonic cells activate more replication origins than differentiated cells. When mouse fibroblast nuclei are introduced into Xenopus laevis mitotic egg extracts, they reprogram to pluripotency and reactivate origins.

Sabarinadh Chilaka (Institut Jacques Monod, Paris, France) used DT40 chicken cells and targeted insertions into a naturally late-replicating region to examine which elements are needed to reschedule replication to the early S-phase. Since active transcription correlates with early replication, an active gene was inserted either alone or in combination with a strong early-origin element, demonstrating no or only mild effects on replication timing, respectively. A better shift into the first half of the S-phase was obtained when the gene and the origin were surrounded by two insulator elements. Chilaka claimed that this structure forms the basic unit of early replication domains [6].

Karl Ekwall (Karolinska Institute, Huddinge, Sweden) concentrated on the insulator function in separating adjacent domains of different chromatin structures. He demonstrated that the chromatin remodeler Fft3 inhibits euchromatin formation in pericentromeric and subtelomeric regions. The deletion of Fft3 in yeast leads to euchromatin invasion into centromeres and subtelomeres, causing misregulation of gene expression and chromosome segregation defects [7].

Rosemary Wilson (University of York, UK) focused on the impact of Cyclin E and Ciz1 on higher-order chromatin structure. Wilson demonstrated that the size of chromatin loops that emanate from attachment points on the nuclear matrix (NM) is reduced in quiescent cells compared with cycling cells. The remodeling of such chromatin loops following re-entry into the cell cycle depends on Cyclin E and Ciz1 positioning at the NM. As both proteins are required for prereplication complex assembly, Wilson concluded that Cyclin E and Ciz1 play a role in remodeling higher-order chromatin structure through the recruitment of active replication origins to the NM.

The implication of nuclear positioning in replication stress induced by DNA breakage was addressed by Marco Foiani (IFOM, Milan, Italy). Using the insertion of a fluorescent inducible gene locus, Foiani and coworkers visualized the gene‘s localization to the nuclear periphery in its active, and to the intranuclear space in its silent form. Stalled forks were detected at the locus upon transcriptional induction when yeast rad53 checkpoint mutant yeast were grown under slowed fork progression. Fork collapsing can be prevented either by creating an additional mutation targeting the transcription unit and nuclear pore-associated protein, TREX-2, or by introducing a double-strand break between the replication fork and the transcribed unit. The combined results suggest that fork stalling at transcribed regions is caused by the gene‘s localization to the nuclear periphery, thereby generating a topological barrier for the fork [8].

Two speakers analyzed the effect of histone dosage on replication reliability and DNA stability. Jakob Mejlvang (BRIC, Copenhagen, Denmark) used micrococcal nuclease experiments and electron microscopy to demonstrate that histone shortage leads to slowed replication and reduced nucleosome occupancy at the daughter strands, while replication forks do not expose ssDNA and remain stable.

Akash Gunjan (Florida State University, FL, USA) and his colleagues overexpressed the histones H3 and H4. This did not increase spontaneous DNA breakage but rendered yeast cells sensitive to DNA damaging agents such as UV radiation, hydroxyurea or bleomycin. Histone reduction on the other hand, leads to increased resistance to these agents. As an elevated recruitment of homologous recombination factors is observed upon decreased histone levels, Gunjan claimed that high histone levels may compete with factors of homologous recombination on the chromatin [9].

A key aspect of the meeting was to answer how replication progresses through chromatin and how the epigenetic state will be reconstituted behind the fork. Duncan Smith (Sloan-Kettering Institute, NY, USA) and his colleagues developed a new technique to purify genome-wide Okazaki fragments from yeast cells using a combination of biochemical approaches and deep sequencing. They demonstrated that the size and distribution of Okazaki fragments are dictated by the position of nucleosomes.

Sara Buonomo (EMBL Mouse Biology Unit, Monterotondo, Italy) described Rif1 as a member of the heterochromatin-specific replication machinery, acting specifically in mid S-phase. Knockdown of Rif1 in mouse embryonic myoblasts impeded early-to-mid S-phase transition [10].

Takaharu Kanno (Karolinska Institute, Stockholm, Sweden) proofed the helix-winding function of the Smc5/6 complex by in vitro experiments. Yeast Smc5/6 mutants demonstrate delayed replication specifically on long chromosomes. These data together suggest that Smc5/6 functions in DNA winding behind replication forks [11].

Constance Alabert (BRIC, Copenhagen, Denmark) and her colleagues developed a new approach called ‘nascent chromatin capture‘, which allows the identification of protein complexes that act during chromatin maturation. Using this method, nascent chromatin was found to be enriched in factors of the replication fork and in histone 4 lysine 5 and 12 acetylation. Histone methylation on newly replicated DNA appears to be a complex process since DNMT2 was detected on the nascent chromatin, whereas another methyltransferase, DNMT3, locates to the mature strand.

Behind the replication fork, newly incorporated histones are acetylated and need to be deacetylated for heterochromation construction. Patrick Varga-Weisz (Babraham Institute, Cambridge, UK) identified SMARCAD1 as the potential histone deacetylase. SMARCAD1 knockdown causes an increase in global H3/H4 acetylation levels, as detected by immunofluorescence. Moreover, upon SMARCAD1 knockdown levels of histone marks that are specific for transcriptional silencing, such as K9me3, decrease, H1 dissociates from the chromatin and chromosome segregation defects accumulate. The combination of these observations supports the hypothesis that SMARCAD1 is implicated in heterochromatin maintenance [12].

Karim Labib (Paterson Institute for Cancer Research, Manchester, UK) presented MCM-2 and FACT as two factors of the replication fork, which are involved in histone recycling. When postreplication cells are DNase treated prior to histone immunoprecipitation, MCM-2 and FACT are isolated together with the histones that have been released from the fork.

Overall, the symposium underlined the functional link between epigenetics, chromatin architecture, transcriptional programming, replication initiation, replication timing and DNA repair. Future research will help us better understand how these mechanisms influence each other.

Acknowledgements

Many thanks to the organizers of the conference: Catherine M Green from the University of Cambridge (UK), Anja Groth from the Biotech Research and Innovation Centre (BRIC) in Copenhagen, Denmark, Camilla Sjögren from the Karolinska Institute in Stockholm and the Abcam Biotech company (Cambridge, UK). The author also thanks all speakers and poster presenters for their contribution to the meeting.

Financial & competing interests disclosure

This work was funded by a postdoctoral fellowship from La Fondation pour la Recherche Médicale (Paris, France). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

This work was funded by a postdoctoral fellowship from La Fondation pour la Recherche Médicale (Paris, France). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.