Epigenetic Regulation in Plants
In the eukaryotic genomes, a DNA molecule is packed inside the nucleosome structure consisting of a core histone protein complex surrounded by approximately 147 base pairs of double strand DNA (Luger et al. 1997). The function of nucleosome is not limited only for storing DNA molecules inside the nucleus. Nucleosome occupancy also takes crucial roles in controlling gene expression (Lauria and Rossi 2011). An In vitro study using mammalian transcription machinery revealed the unsuccessful transcription initiation in the promoter of the gene which contacts with the nucleosome structure (Lorch et al. 1987). In contrast, the depletion of single histone subunit leading to the incomplete nucleosome increases the global gene expression in yeast (Han and Grunstein 1988). In higher plants such as Arabidopsis and rice (Oryza sativa), there is a negative correlation between the nucleosome occupancy in the promoter region with the expression level of the gene (Zhang et al. 2015). Results from in vitro and in vivo studies suggest the function of nucleosome as an obstacle of gene expression in eukaryotic cells. Thus, to gain the accessibility of DNA sequence for transcription machinery, eukaryotes (including plants) use three systems to alter chromatin structure: DNA methylation, histone modifications, and chromatin remodelling protein complexes (Bell et al. 2011; Lauria and Rossi 2011). These methods are considered as epigenetic marks due to their abilities to regulate the expression of genes without any change in DNA sequence.
1.3.1 DNA Methylation
DNA methylation occurs by transferring methyl group from the donor, S- adenosylmethionine, to the cytosine (C) residue of the target nucleotides (Smith et al. 2010). In plants, there are three patterns of DNA methylation and can be categorised within two groups: the symmetric CG and CHG methylation (where H is A, C, or T) and asymmetric CHH methylation (He et al. 2011) and there are two processes of DNA methylation in plants: (1) de novo DNA methylation (the establishment of DNA methylation) and (2) the maintenance of DNA methylation pattern.
To establish all DNA methylation patterns in plant genome, plants use RNA-directed DNA methylation (RdDM). As shown in figure 3A, the process of RdDM consists of several steps. Firstly, the production of double strands 24-nt small interfering RNAs (siRNAs) that is carried out by RNA polymerase IV (Pol IV), RNA-DEPENDENT RNA POLYMERASE 2 (RDR2), and DICER-LIKE 3 (DCL3). Then, ARGONAUTE 4 (AGO4) involves in the process by carrying only one strand of siRNA and form the protein complex with RNA polymerase V (Pol V). Finally, the protein complex of AGO4 and Pol V directs DOMAIN REARRANGED METHYLTRANSFERRASE 2 (DRM2, the methyl transferase enzyme) to transfer the methyl group to specific DNA sequences. Apart from the establishment of all methylation pattern, each methylation pattern is maintained by different process. For example, in CG methylation, plants use DNA methyltransferase 1 (MET1) together with VARIANT IN METHLATION1 (VIM1) to methylate the newly synthesised DNA during DNA replication (figure 3B, Hauser et al. 2011; Kawashima and Berger 2014; Law and Jacobsen 2010). To maintain CHG methylation, DNA methyltransferase specific to plant named CHROMOMETHYLASE 3 (CMT3) is used together with the histone methyltransferase SU(VAR)3-9 HOMOLOG4/KRYPTONITE (SUVH4/KYP). SUVH4 transfers two methyl groups to the 9th lysine residue in histone 3 (H3) subunit generating H3K9me2 which is required for the function of CMT3 (Figure 3C, (Hauser et al. 2011; Kawashima and Berger 2014; Stroud et al. 2014). Because plant cells are unable to copy asymmetrical methylation pattern directly to the new strand DNA during the cell division, the maintenance of CHH methylation is carried out by the process of re-establishment (Smith et al. 2010). CHH can be re-established by the RdDM pathway using 24nt siRNAs mediated by SAWADEE HOMEODOMAIN HOMOLOGUE 1 (SHH1) binding with H3K9me2 or by RdDM pathway using 21-22nt siRNAs produced from RNA polymerase II (Pol II), RNA-DEPENDENT
RNA POLYMERASE 6 (RDR 6), and AGONAUTE 2 (AGO2) (figure 3D, Kawashima and Berger 2014; Matzke and Mosher 2014).
In contrast to the maintenance, the methyl group(s) of methylated cytosine are removed by either a loss of maintenance during DNA replication or from the activity of DNA glycosylase (Law and Jacobsen 2010). In plants, demethylation is carried out by DNA glycosylases: DEMETER (DME) or REPRESSOR OF SILENCING 1 (ROS1) (Choi et al. 2002; Gong et al. 2002). Report from Baute and Depicker (2008) proposed that base excision repair is the possible mechanism to demethylate the methyl group due to the functions of DME and ROS1 that can remove nitrogenous base (i.e. methylated cytosine) and break the backbone of DNA. Although DME and ROS1 have the same function in DNA demethylation, activities of DME and ROS1 are different in plant developmental stages. While ROS1 acts in vegetative stage, DME has an ability to demethylate DNA during gametogenesis (Law and Jacobsen 2010).
All methylation patterns are mainly found in the area of transposable elements and repetitive DNA sequence in Arabidopsis genomes (Zhang et al. 2006; Zilberman et al. 2007). Only 5% of DNA methylation in Arabidopsis is found in gene promoters and the methylation in promoter regions shows higher effect on the expression more than the methylation in transcribed regions (Zhang et al. 2006).
homodomain homologue 1 (SHH1) binds to H3K9me2 and recruits RNA Pol IV. RNA pol IV, RDR2, and other components in RdDM pathway generate 24-nt siRNAs which bind to AGO4 to activate DRM proteins (DRMs) to establish DNA methylation in all contexts. Non-canonical RdDM includes activities of RNA polymerase II and RDR6 to generate 21-nt siRNAs to activate DRMs. (Modified from: Kawashima and Berger 2014).
1.3.2 Histone Modification
Histone core in nucleosome consists of 8 subunits (two of each H2A, H2B, H3, and H4). Each histone has the extension of amino acid residues at N-terminal which are important in the alteration in chromatin structure (Albert et al., 2002). In plants, four types of post- translational modification of histone are observed to have effects on gene transcription: ubiquitination on lysine residues (showing either positive or negative effect on transcription), methylation on lysine and/or arginine residues (showing either positive or negative effect on transcription), acetylation on lysine residues (causing transcription activation), and phosphorylation on serine and/or threonine residues (causing transcription activation) (Pfluger and Wagner 2007). The other mechanisms include sumoylation and ADP ribosylation (Lauria and Rossi 2011). Histone acetylation and methylation are the most well-studied mechanisms among all histone modifications especially at N-terminal tails of histone H3 and H4 subunits that are understood to affect gene regulation (Hauser et al. 2011).
Histone acetylation and deacetylation are carried out by two enzymes: histone acetyltransferase (HAT) and histone deacetylase (HDAC), respectively. The acetylation on lysine residues decreases positive charges of histone subunits leading to less compacted nucleosome structure that can open DNA region and activate gene transcription. Thus, histone acetylation is considered to be the positive mark of transcription while histone deacetylation by HDAC gives the opposite results (Shahbazian and Grunstein 2007). Many of histone acetylation marks and their effects are observed in Arabidopsis: acetylations at the 5th, 8th, 12th and 16th of N-terminal lysine residues of histone H4 together with H3K18Ac (an acetylation at 18th lysine of H3 subunit) are involved in changes of chromosome structure during cell cycle. Importantly, an acetylation at H3K9 is found to be involved with highly transcribed genes at euchromatin regions (Boycheva et al., 2014).
Histone methylation is carried out by histone lysine methyltransferase (HMKT). Unlike histone acetylation that always cause the transcription activation, histone methylation establishes variant effect depending on the location and degree of methylation (mono-, di-, or tri-methylation) (Hauser et al. 2011; Liu et al. 2010). Histone methylation in lysine residue is also reversible by two different demethylase enzymes: lysine-specific demethylase I (LSD1) removing methyl group only from mono- and di-methylated lysine and Jumonji C (JmjC) which is able to demethylate from all degrees (Hauser et al. 2011). The examples of histone methylation in Arabidopsis include the heterochromatin marks (causing inactivation of
transcription) such as H3K9me1 (mono-methylation at lysine 9 of H3 subunit), H3K9me2 (di- methylation at lysine 9 of H3 subunit), and H3K27me1 (Liu et al. 2010). H3K9me2 is also involved in the maintenance of CHG methylation in DNA (Hauser et al. 2011; Kawashima and Berger 2014; Stroud et al. 2014).
1.3.3 ATP-dependent chromatin remodelling complexes
While DNA methylation and histone modification focus on adding or removing functional groups at DNA or histone molecules, the modulation by chromatin remodelling complexes (CRCs) alters the interaction between DNA and histone without changing the chemical components. This process is done by using protein complexes such as SWITCHING DEFECTIVE2/SUCROSE NONFERMENTING2 (SWI2/SNF2) family (Clapier and Cairns 2009; Peterson et al. 1994). In eukaryotes, SWI2/SNF2 family can be classified into 4 groups due to the protein sequence: SWI/SNF subfamily, Imitation Switch (ISWI) subfamily, Chromodomain Helicase DNA-Binding (CHD) subfamily, and Inositol requiring 80 (INO) subfamily (Clapier and Cairns 2009; Ojolo et al. 2018). Each of them takes crucial roles in different biological aspects in eukaryotic cells. However, all chromatin remodellers contain ATPase/helicase domain to utilise energy from ATP hydrolysis for modulating the interaction between DNA and histone (Clapier and Cairns 2009; Ojolo et al. 2018). The activity of CRCs results in nucleosome repositioning, ejection, unwrapping, or composition altering by histone variants leading to accessible (open) DNA regions that can interact with DNA-binding domains of proteins involved in transcription, DNA replication, or DNA repair (Clapier and Cairns 2009). In plants, CRCs are involved in all stages in life cycle of each plant including the maintenance of shoot and root apical meristem (SAM and RAM), cell differentiation, organ differentiation, and the development of reproductive organs (Ojolo et al. 2018). Moreover, one of CRCs named BRAHMA (BRM) shows its physical interaction with the important components in abscisic acid (ABA) pathway, the phytohormone used in stress response in plants (Peirats-Llobet et al. 2016).
In conclusion, DNA methylation, histone modification, and chromatin remodelling are the mechanisms used by plants and other eukaryotes for controlling an interaction between histone and adjacent DNA. These mechanisms are carried out by enzymatic processes or protein complexes and all of them are reversible which allow plant to use them in time- specific or cell type-specific manners in the development.