Project Topics
Get Complete Project Material File(s) Now! »
Histone H3 variant CENP-A as the determinant of a functional centromere
Location of the active centromere is epigenetically determined by the histone H3 variant CENP-A that assembles at the centromeres of eukaryotic chromosomes. CENP-A is the only known signature of a functional centromere and is also found on neocentromeres formed on non-repetitive DNA. CENP-A homologues have been found in all eukaryotes, and are known under different names such as CENP-A in mammals, CID in Drosophila, Cse4 in S.cerevisiae. Downregulation of CENP-A is lethal for all organisms, causing defects in chromosome segregation. The loss of CENP-A results in mislocalization of inner kinetochore proteins and failure of kinetochore assembly leading to aneuploidy and genomic instability (Howman et al. 2000; Régnier et al. 2005). Aneuploidy is a mark of almost all tumours and it is most likely caused by CENP-A overexpression and the formation of ectopic neocentromeres as observed in human colorectal cancer. CENP-A overexpression in colorectal cancer cells leads to CENP-A mislocalization to noncentromeric regions of chromosome or even to a complete loss of CENP-A, suggesting disruption of the kinetochore (Tomonaga et al. 2003). Indeed, abolishment of CENP-A overexpression decreases aneuploidy (Amato et al. 2009). If CENP-A overexpression is indeed the main reason for chromosomal instability in cancer cells, understanding the mechanisms leading to this event and its consequences might help in developing new approaches in anticancer therapy. How CENP-A incorporation beyond the centromere boundaries is limited is still not known but it might be that the formation of pericentromeric heterochromatin prevents further CENP-A incorporation (Pidoux & Allshire 2005).
CENP-A containing nucleosomes carry structural features that distinguish them from the canonical H3-containing nucleosomes. Amino acid sequence of the N-terminal tail of CENP-A differs from histone H3 and is also quite variable between species. However, only the region localized at the C-terminal histone fold domain called the CENP-A targeting domain (CATD) is required to target CENP-A to chromosomes (Black et al. 2004). Histone H3 containing the 22 amino acids that make the CATD domain is able to replace CENP-A at the functional centromere (Black, Jansen, et al. 2007). Due to the CATD domain, the CENPA nucleosomes show a more rigid structure to the interface formed with the histone H4 (tenfold more slower hydrogen exchange along the peptide backbone) than the histone H3 (Black, Brock, et al. 2007).
The canonical histone H3 contains numerous post-translational modifications on its Nterminal tail, while only some are known for CENP-A (Bailey et al. 2013). CENP-A is phosphorylated by Aurora B at serine 7 (Ser7), which is similar to Aurora B phosphorylation of histone H3 at Ser 10 (Zeitlin et al. 2001). Other modifications characteristic only to CENPA are trimethylation of glycine 1 (Gly1) and phosphorylation of Ser16 and Ser18. The double serine phosphorylation motif was shown to form a specific structure that cause intramolecular associations between the N-terminal tails of CENP-A, changing the conformation of CENP-A nucleosomes and resulting in a different chromatin structure at the centromere (Bailey et al. 2013).
After each round of replication, centromere site must be re-established at the newly synthesised sister chromatid. The CENP-A nucleosomes are loaded at the place of the preexisting CENP-A by the help of a histone chaperone HJURP (Holliday junction recognition protein) during the early G1 phase (Dunleavy et al. 2005; Foltz et al. 2009).
Centromere associated proteins
The kinetochore is seen by electron microscopy as a three-laminar structure at the primary constriction of chromosomes. It is composed of distinct protein complexes attached on one side to the microtubules and on the other side to the centromeric chromatin, allowing the segregation of sister chromatids during cell division (Cheeseman & Desai 2008) (Figure 8). The outer kinetochore layer consists out of several protein complexes. These complexes together form the KMN network of proteins, named according to the acronym for the protein components KNL1, Mis12 and Ndc80, that bind to the microtubules (Cheeseman et al. 2004; Cheeseman et al. 2006). KMN network interacts with proteins of the inner kinetochore together called the CCAN (constitutively centromere-associated network). The CCAN is formed by a group of centromere specific proteins termed CENPs (for centromeric proteins).
There are 16 CENPs that permanently associate with the centromere (Foltz et al. 2006; Okada et al. 2006). The localisation of CCAN occurs downstream to CENP-A deposition suggesting that CENP-A marks the position for the CCAN assembly. After the initial recruitment to the centromere, the CCAN acts as a foundation for the assembly of the outer kinetochore proteins (Hori et al. 2008; Hori et al. 2013). CENP-C localisation to the centromere requires six Cterminal residues of the CENP-A (Guse et al. 2011). The only centromeric protein that specifically binds to a centromeric sequence is CENP-B. CENP-B binds to a specific 17 bp sequence present at the centromeric region of human and mouse, and it seems to be required for de novo assembly of CENP-A (Masumoto et al. 1989). Chromosomal passenger complex containing Aurora B kinase, INCENP, Survivin and Borealin transiently mark the centromere during mitosis (Ruchaud et al. 2007).
Centrochromatin – the chromatin forming at the centromeres
Defined by the presence of CENP-A, centromeric chromatin reveals a specific level of organisation and structure. On the extended chromatin fibre, chromatin at the centromere core in both flies and humans is arranged into regions containing CENP-A/CID nucleosomes interspersed with regions containing the canonical histone H3 nucleosomes (Figure 9). H3 CENP-A adjacent along the DNA fibre but rather scattered between the blocks of H3 containing nucleosomes. Core histones were detected by immunofluorescence on extended chromatin fibres (Black & Bassett 2008). 3D organisation of the metaphase chromosomes shows the formation of a unique cylindrical structure where blocks of CENP-A/CID nucleosomes are orientated toward the outer kinetochore plate while H3 containing nucleosomes are placed toward the interior, between the sister chromatids (Blower et al. 2002) (Figure 10). The length of the DNA forming this structure is approximately 500-1500 kb in humans and 200-500 kb in flies (Allshire & Karpen 2009). This distinct chromosomal domain sometimes referred to as centrochromatin is surrounded by long stretches of heterochromatin. These two regions are marked by specific epigenetic marks that distinguishes them one from the other and from the rest of the genome. The core histone H3 at the centromere is uniquely modified carrying posttranslational modifications different from both silent and active chromatin. Centromeric H3 is dimethylated at lysine 4 (H3K4me2) which is an epigenetic mark associated with euchromatin
and potentially transcriptionally active regions. However, both histones H3 and H4 of centromeric nucleosomes are hypoacetylated, lacking acetylation marks usually found in euchromatin, and are at the same time deprived of di- or tri-methylation of H3K9, a hallmark of heterochromatin. Heterochromatic regions that flank the centromeric chromatin are enriched for H3K9 di- and trimethylation and show hypoacetylation of both H3 and H4 histones (Sullivan & Karpen 2004). The reason for the formation of this unique type of chromatin on the centromere is still questioned. The existence of canonical histone H3 nucleosomes at the centromere, carrying distinct epigenetic marks surely has a functional significance. It participates to the formation of the three-dimensional structure on the metaphase chromosome, assuring kinetochore assembly and contacts with the microtubules. Canonical histone carrying the lysine 4 dimethylation mark seems to promote the incorporation of CENP-A in human cells. Indeed, depletion of H3K4me2 at centromeres fails to recruit CENP-A chaperon HJURP causing defectiveness of CENP-A incorporation (Blower et al. 2002, Bergmann et al. 2011). The combination of histones together with specific epigenetic marks could promote the incorporation of CENP-A to the locus, marking it for kinetochore assembly in the next generations (Allshire & Karpen 2009). These findings confirm the importance of CENP-A as the foundation for the assembly of kinetochore and centromere function. These structural properties make CENP-A nucleosomes different from the canonical H3 nucleosomes and surely facilitate the assembly of a specific chromatin on the centromere.
Epigenetic signature of pericentromeric heterochromatin
Pericentromeric heterochromatin forms at repetitive, AT-rich satellite DNA sequences and carries a specific pattern of epigenetic modifications. With the exception of S. pombe, DNA at heterochromatin is highly methylated. Histones are generally hypoacetylated and carry a characteristic methylation pattern: they are enriched in histone 3 lysine 9 di- and trimethylation (H3K9me2, H3K9me3), histone 3 lysine 27 monomethylation (H3K27me1) and histone 4 lysine 20 trimethylation (H4K20me3) marks (Peters et al. 2001; Kourmouli et al. 2004; Martens et al. 2005). Non-histone proteins such as HP1 associate with pericentromeric heterochromatin.
HP1 was first discovered as a suppressor of position effect variegation (PEV, see paragraph 2.3.4 of this Chapter), when a mutation in Su(var)2-5, a gene coding for HP1, resulted in inhibition of PEV (Eissenberg et al. 1990). Three isoforms of HP1 are known in mammals: HP1α, HP1β and HP1γ. These proteins are evolutionary conserved, and can be found throughout eukaryotic organisms. HP1 homologs in S. pombe are known under the names Swi6 and Chp2 (Zeng et al. 2010). Mammalian HP1α and HP1β localize to heterochromatic regions, while HP1γ can be found in euchromatin, where it is implicated in transcriptional activation (Minc et al. 2000).
Another non-histone component of pericentromeric heterochromatin is SU(VAR)3-9, a histone methyltransferase (HMT) conserved from mammals (Suv39h1 and 2) to fission yeast (Clr4). Suv39h is primarily responsible for trimethylation of H3K9 (Rea et al. 2000; Rice et al. 2003). Other methyltransferases such as SETDB1 and G9a also modify H3K9 in mammals, catalysing different levels of methylation (mono- and dimethylation). Suv39h indirectly controls DNA methylation and methylation of H4K20. Loss of Suv39h activity causes absence of H3K9 methylation mark and loss of HP1 from heterochromatin (Bannister et al. 2001; Peters et al. 2001).
HP1 contains three distinct domains that have different binding preferences (Figure 11). HP1 chromodomain present in its N-terminal end tethers HP1 to heterochromatin through the interaction with Suv39h-trimethylated lysine 9 of histone H3 (Lachner et al. 2001a; Bannister et al. 2001; Nakayama et al. 2001). This highly specific interaction cannot be seen in other chromodomain proteins such as Polycomb. Indeed, H3K9me3 binds a conserved motif present in HP1 chromodomain. However, since H3K9me3 is not specific only to heterochromatin, but appears throughout the nucleus, this interaction seems not to be sufficient for HP1 targeting to heterochromatin. Moreover, the central, variable hinge domain of HP1 is found to strongly bind RNA (Muchardt et al. 2002). HP1 hinge domain is a major target to sumoylation. It was shown that this post-translational modification promotes association of HP1 with RNA and its accumulation to pericentromeric heterochromatin (Maison et al. 2011). H3K9me3 provides binding sites for HP1 that in turn binds Suv39h enzyme with its chromoshadow domain (Lachner et al. 2001a; Bannister et al. 2001; Nakayama et al. 2001). HP1 chromoshadow domain can also bind other proteins such as Suv4-20h2, methyl-CpG-binding proteins, Dnmt3a and Dnmt3b and the large subunit of chromatin assembly factor CAF, p150, implicated in DNA replication and repair. It also interacts with HP1α and HP1β, a feature that permits homodimerisation of HP1 proteins.
Table of contents :
Abbreviations
Table of contents
Figures
Tables
Introduction
CHAPTER I Structure and function of centromeric regions Usage of the terms
I. 1. Determining the centromeric region
1.1. First description of the centromere
1. 2. Organisation of the centromeric region
1.3. DNA at the centromeric and pericentromeric region
1.3.1 Repetitive DNA
1.3.2. Diverse functions of repetitive DNA
1.3.3. Repetitive DNA at the centromeric region
1.3.4. Variation of centromeric sequences between species
1.4. Neocentromeres and dicentrics
I. 2. Chromatin organisation at the centromere
2.1. Histone modifications and the underlying chromatin state
2.2. The centromere core
2.2.1. Histone H3 variant CENP-A as the determinant of a functional centromere
2.2.2. Centromere associated proteins
2.2.3. Centrochromatin – the chromatin forming at the centromeres
2.3. Pericentromeric region
2.3.1. Epigenetic signature of pericentromeric heterochromatin
2.3.2. The interaction network which allows the establishment and maintenance of pericentromeric heterochromatin
2.3.3. Heterochromatin assembly in S pombe requires RNA interference
2.3.4. Role of pericentromeric heterochromatin
2.4. Centromeric and pericentromeric regions in mouse cells
2.4.1. Organisation of mouse centromeric and pericentromeric region
2.4.2. Association of pericentromeric regions
CHAPTER II. Non-coding RNA from centromeric and pericentromeric regions
II.1 Non-coding RNA
1.1. The RNA world, old and new
1.1.1. A new perspective on RNA
1.1.2. Non-coding RNA as a key to complexity
1.2. Emerging role for ncRNAs
1.2.1. Short ncRNA
1.2.2. Long ncRNA
II.2. Expression of repetitive sequences
2.1. Evidence for transcription from centromeres and pericentromeres
2.2. Centromeric transcripts as integral components of centromeric chromatin
2.2.1. Centromeric RNAs regulate the kinetochore activity
2.2.2. Centromeric transcription stabilise CENP-C binding to the centromeres
2.3. Transcription from the pericentromeric region
2.3.1. Pericentromeric transcription during stress
2.3.2. Pericentromeric transcripts participate to heterochromatin reorganisation during development and differentiation
2.3.3 Non-coding RNA in heterochromatin formation: lessons from fission yeast .
2.3.4. Non-coding RNA as a component of pericentromeric heterochromatin
II.3. Regulation of pericentromeric transcription
3.1. Chromatin modifications and transcription
3.1.1.Histone modifications
3.1.2. DNA methylation
3.2. Transcription factors
3.3. Cell cycle
Chapter III Tools for study of repetitive sequences
I.1. Oligonucleotides for detection of nucleic acids
1.1. Hybridization properties of nucleic acids
1.2. Oligonucleotides
1.2.1 A brief history of oligonucleotides
1.2.2. Oligos with chemical modifications: 2’-O-Me and LNA
1.3. Locked nucleic acids
1.3.1 Thermodynamic properties of LNA
1.3.2. Use of LNA oligonucleotides
1.4. LNA probes for the study of repetitive sequences
I. 2. Tools for (epi)genetic engineering
2.1. Tools for targeted genome manipulation
2.1.1. Targeting specific DNA loci in living cells
2.1.2. TALE as a novel DNA binding domain
2.2.3. Designer TALEs
2.2. Epigenetic engineering for studying the functions of chromatin modifications
2.2.1. Targeting chromatin modifications in living cells
Objectives
Materials and methods
I.1. Methods
1.1. Cell culture
1.2. Northern blot
1.3. Radioactive labelling
1.4. RNA extraction
1.5. RT PCR
1.6. Major satellite DNA probe preparation
1.7. In vitro transcription
1.8. Small RNA separation
1.9. Cell transfection
1.10. Immunofluorescence
1.12. Microscopy
1.13. TANGO analysis
I.2. Materials
2.1. Oligonucleotides
2.2. Primers
Results
CHAPTER I Characterization of major satellite transcripts
I.I. Characterization of major satellite transcripts using LNA oligonucleotides
1.1. Probes for detection of major satellite repeats
1.2. Characterisation of major satellite transcription by northern blotting of total RNA from mouse cells
1.2.1. Transcriptional profile of major satellites in growing mouse cells
1.2.2. Detailed characterization of northern hybridization signals
1.2.3. Comparison of transcriptional profile revealed by isosequential LNA and 2’- O-Me oligo
1.2.4. High stringency washing
1.2.5. RNA and DNA probes
2.2. Strand specific RT PCR confirms transcription from major satellites
I.2. Expression and regulation mechanisms implicated in major satellite transcription
2.1. Influence of the inhibitors of chromatin modifiers on major satellite transcription
2.2. Changes in major satellite transcription upon thermal stress
2.3. Influence of different RNA polymerase inhibitors on major satellite transcription
I.3. Sequence characterization
3.1. Technical details/methods for major satellite sequence characterization
I. 4. Conclusion and discussion
CHAPTER II TALE fused to histone demethylase mJMJD2D for epigenetic engineering at pericentromeric regions of mouse cells
II.1. Targeting mouse major satellites using TALE fused to histone demethylase mJMJD2D
1.1. Context of the study
1.2. Visualisation of the specificity of TALE recruitment on major satellites
1.3. Loss of H3K9me3 upon transfection with TALE fused to histone demethylase mJMJD2D
II.2. Analysis of the effect of histone demethylation upon transfection with TALEs fused to histone demethylase mJMJD2D
2.1. Tools for Analysis of Nuclear Genome Organisation (TANGO)
2.2. Quantitative analysis of the demethylation of H3K9me3 at the major satellite foci upon transfection with the TALE N212-mJMJD2D
2.3. Effect of the demethylation of H3K9me3 on major satellite foci
II. 3. Conclusion and discussion
General conclusion and perspectives
Appendix
Bibliography
