1. Plant Epigenetic Regulation Mechanism
The term ‘epigenetics’ was first used by Conrad Waddington in 1942
[1][33] to describe the interactions between genes and the environment, which, in consequence, was to lead to the development of certain phenotypes
[2][30]. Considering the molecular aspect, the term ‘epigenetics’ means the study of heritable changes involving changes in gene expression, but not in DNA sequences
[3][4][24,34]. Epigenetic changes are based on molecular processes, such as methylation of cytosine residues in DNA, chromatin remodelling by modifying histone proteins, and regulatory processes mediated by small non-coding RNA
[5][6](Figure 1) [35,36].
DNA methylation is a stable modification consisting of the covalent attachment of a methyl group to the fifth position of the cytosine pyrimidine ring (5 mC) or the sixth position of the adenosine purine ring (6 mA). It is catalysed by DNA methyltransferase using S-adenosyl-L-methionine as a methyl donor
[7][8][9][37,38,39]. This is important for various biological processes, mainly related to gene and transposon silencing. The number and location of methyl residues in the promoter or coding sequence of given genes have large functional consequences for this gene
[8][10][38,40]. Modifications caused by DNA methylation in plants can be passed to the next generation through meiotic divisions or, reversibly, wiped out during mitotic cell division
[11][41]. In plants, the methyl group (-CH3) is attached to the cytosine in three sequential contexts: (1) CG, catalysed by methyltransferase 1 (MET1)
[12][42], (2) CHG (H = A/T/C), maintained by a plant-specific chromomethylase 3 (CMT3), in association with dimethylated lysine 9 in histone H3 (H3K9me2)
[13][43] and (3) CHH, methylated by domain rearrangement methyltransferase (DRM2). DRM2-dependent DNA methylation is driven by small interfering RNAs (siRNAs) via the RNA-dependent DNA methylation (RdDM) pathway with consequent gene silencing
[7][8][10][37,38,40]. Key factors related to these epigenetic modifications have been identified. There are Writers, which are enzymes responsible for the modification of nucleotide bases in DNA and amino acid residues in histone proteins
[14][44]. The next group are Erasers, which are enzymes that erase changes established by Writers. The last group of factors are Readers, which are proteins with specific domains binding or interacting with epigenetic signs located in the locus
[14][44].
Modifications of histones affect modifications of the chromatin structure in response to endogenous stimuli and changes in the environment
[15][17]. Modifications of histone proteins include methylation, acetylation, phosphorylation, ubiquitination, and sumoylation. All of the above-mentioned modifications influence the formation of chromatin states
[6][36]. One of the best-known histone modifications is methylation. Both lysine (K) and arginine (R) can undergo mono-, di-, or trimethylation (me1, me2, and me3)
[16][45]. Histone methylation is mainly mediated by proteins containing the SET (Su(var)3–9, Enhancer-of-zeste, and Trithorax) domain. Modified histones are recognized by the corresponding proteins, which, together with other ATP-dependent remodelling complexes, make further changes in the availability of genetic information
[17][46]. Highly conserved proteins from the Trithorax (TrxG) and Polycomb (PcG) families play a major role in regulating the expression of genes influencing developmental states in living organisms
[18][47]. They also maintain the memory of the transcriptionally active or inactive chromatin status during stress
[19][48]. TrxG and PcG play opposite roles in regulating gene expression of defence responses. It is generally accepted that the TrxG-mediated methylation markers associated with transcriptional activation, H3K4me2/3, maintain the trainable genes in a transcriptionally active state
[20][49]. Thus, TrxG exhibits an antagonistic activity against the Polycomb family proteins which establish H3K27me3 and H3K9me2 methylation with a rather repressive effect on gene expression
[21][22][50,51]. The chromatin structure may be regulated by histone methyltransferases and demethylases by the control of all degrees of lysine methylation, thus regulating various functions in the cell
[23][24][25][52,53,54].
Histone acetylation was first described in 1964 by the Allfrey team
[26][55]. The acetylation and deacetylation of the histone tails are mediated by enzymes known as acetylases (HATs) and histone deacetylases (HDACs), respectively. Many studies have found that HAT plays an important role in plant development and stress response
[27][56]. Histone deacetylases (HDACs) are composed of three different families, of which HD2 are plant-specific deacetylases. Histone deacetylases lead to gene silencing through chromatin condensation and removal of the acetyl group from histone proteins
[28][57]. Another modification of histones is phosphorylation. This modification is crucial for gene transcription activation, DNA repair, cell cycle-dependent chromosome condensation and segregation, as well as apoptosis
[29][30][58,59]. Both serines (S) and threonines (T) are phosphorylated
[31][60]. Another histone modification, ubiquitination, involves the attachment of a small conserved protein composed of 75 amino acids—ubiquitin to histone proteins. The linker histones and those that make up the nucleosomes are ubiquitinated
[32][61]. Ubiquitin is attached to the ε-amino group of selected lysines by an isopeptide bond. This is a reversible phenomenon. Ubiquitination also serves as a signal for protein degradation that occurs in the proteasomes. It may also affect the subcellular localization and biochemical activity of the target protein
[32][33][61,62]. Sumoylation is a modification of histones, in which the SUMO (small ubiquitin-related modifier) protein takes part. Like ubiquitin, the SUMO protein can covalently associate with other proteins using specific enzymes. Sumoylation is associated with the regulation of transcription and the stress response of plants
[34][35][63,64].
Important in understanding the impact of chromatin modifications on gene regulation was the discovery that, even though different classes of epigenetic modifications act independently of each other, they can very often influence the recruitment of complex protein complexes regulating the transcriptional activity of genes in a complex manner
[5][6][35,36]. The discovery of the importance of individual histone modifications led to the creation of a histone code that carries information about the activity of a given area of the genome. The histone code hypothesis assumes that a given modification of a non-specific histone residue determines modifications to the same or a different histone. Moreover, individual modifications or their combinations are “read” by protein complexes remodelling the chromatin structure, which influences the transcription of genes
[36][37][65,66]. Chromatin remodelling complexes often use ATP to perform histone–DNA interactions, which, in turn, lead to changes in the nucleosomes. Proteins of the SWI/SNF class form the main complexes remodelling chromatin. The presence of DNA binding domains indicates the possibility of recruitment of these complexes by transcription factors
[38][67]. In addition, these complexes play a key role in regulating growth and maintaining plant-specific dynamics of developmental changes. Remodelling complexes also interact with histone-modifying proteins, including from the TrxG and PcG family through bromo- or chromodomains
[39][68]. In turn, the SWR complex is responsible for the exchange of histones in nucleosomes and the activation of gene transcription
[31][60]. Regardless of whether the end effect is the activation or repression of transcription, this process is carried out through the complex action of complexes affecting histone proteins and transcription factors that act comprehensively
[39][68].
RNA-directed DNA methylation (RdDM) is a biological process in which small, non-coding RNA molecules are involved in driving changes in DNA methylation through specialized transcription machinery
[40][41][69,70]. RdDM depends on specialized transcription machinery clustered around two plant polymerases—Pol IV and Pol V
[42][71]. In a process supervised by Pol IV, small interfering RNAs (siRNAs) are generated and transported to the cytoplasm. There they are attached to the AGO4 protein (ARGONAUTE 4) and re-transferred to the cell nucleus where siRNA directs the recruitment of Pol V transcripts and the targeting of DRM2 methyltransferase activity, consequently mediating de novo cytosine methylation in all sequence context classes (i.e., CG, CHG, and CHH)
[40][41][43][44][69,70,72,73].
MicroRNAs (miRNAs) are small, endogenous, non-coding RNAs, about 20–25 nucleotides long, which are designed to suppress target gene expression through sequence complementarity
[45][74]. miRNAs are involved in the inhibition of gene expression by directing the RNA-induced silencing complex (RISC) to mRNA on the basis of base complementarity
[46][75]. By creating the RISC complex, the miRNA is attached to the AGO1 protein, consequently causing mRNA silencing through targeted cleavage or translation inhibition
[47][48][76,77].
2. Epigenetics in Tree Development
Forest trees are perennial organisms characterised by complex life cycles that are exposed to changing environmental conditions during their long lifespans
[11][41]. From the end of the 20th century onward, Europe has been exposed to an increase in temperature combined with a deficit in rainfall, which has negatively impacted both the health and vitality of forest stands; this, in turn, may lead to significant social and economic losses
[49][50][80,81] as forests play an essential role in maintaining the environmental balance, storing CO
2, preventing soil erosion, supplying wood, etc.
[51][82]. The adaptation strategies of forests to changing climatic conditions, including high temperatures and drought, is the paramount importance for the preservation of such properties
[52][53][83,84].
Climate change is influencing the availability of resources and conditions that are critical to plant performance. Currently, some tree species are considered resistant to the effects of climate change
[49][80]. One of the ways plants respond to changing environmental conditions is through the acclimatization by phenotypic plasticity
[54][55][6,85]. Phenotypic plasticity is manifested through the ability of a single genotype to produce different phenotypes depending on the environment
[56][86]. Recent studies provide important evidence that epigenetic mechanisms at the base of phenotypic plasticity are essential for stress responses
[57][58][59][87,88,89] and can enable organisms to response rapidly to environmental changes, including climate change
[60][61][90,91]. The loss or maintenance of epigenetic changes associated with changes in the environment of a particular plant enables a fair start for a new generation and ensures the growth and development of the offspring at the same level
[62][63][92,93].
Silva et al.
[64][94] investigating in silico DNA methyltransferases, DNA demethylases, and other histone modifiers in
Quercus suber showed a link between the expression levels of each gene in different tissues (buds, flowers, acorns, embryos, cork, and roots) with the functions already known. They imply that the data generated during such investigation may be important for future studies exploring the role of epigenetic regulators in this economically important species.
3. Dormancy and Germination of the Seeds
Research conducted by Reich and Oleksyn
[65][114] indicated that the obtained from
Pinus sylvestris trees had a specific memory of the climatic conditions in which their mother plants grew, which, in turn, affected their ability to germination in specific environmental conditions. Yakovlev et al.
[61][91] provided evidence for the occurrence of an epigenetic memory phenomenon responsible for adaptation to variable environments in the common spruce (
Picea abies). They indicated that the temperature fluctuations occurring during embryogenesis altered some characteristics of the adult individuals, such as frost tolerance, bud phenology, and seed production, suggesting a specific embryo-mature memory. Environmental factors during the storage of seeds in soil have a key impact on their later viability and germination which, in turn, can be passed to the next generation
[66][67][115,116].
Alakärppä
[68][117] conducted studies on DNA methylation and expression of selected DNA methyltransferase (DNMT) genes on mature
Pinus silvestris L. seeds from three populations collected in northern and southern Finland. A correlation was found between climatic factors and the expression of DNMT genes in embryos, which may suggest that these changes contribute to the local adaptation of Scots pine. In addition, a variety of DNA methylation levels combined with changes in the expression of the studied genes may contribute to the improvement of the condition of trees in dynamically changing environmental conditions.
Seed dormancy is one of the most important elements of plant performance which delays the germination until optimal environmental conditions are appropriate for further growth and development. It is a complex trait caused by genetic factors and controlled by environmental conditions
[69][118]. After seed dispersion, in soil seed banks, under natural conditions, the level of dormancy usually changes dynamically in the annual cycle, and the beginning of the growing season is associated with the highest seed germination potential. Climate changes (e.g., temperature and precipitation) may affect the durability of seeds stored in soil banks, targeting characteristics, such as their longevity, dormancy depth, and pathogen resistance
[69][118]. The persistence of an exemplary seed population in a given environment depends on its resistance to premature emergence from the seed bank by germination or death and its exposure to the environmental conditions that favour this fate
[70][119]. In this case, geographical distribution and rapid temperature changes may turn out to be unfavourable for a given species, given that the migratory capacity of woody plants affects their reaction in a limiting manner
[69][118]. Changes in plant distribution ranges caused by climate change may not only result in migration to new areas that are more suitable for a given organism but may also select against phenotypes that adapt poorly to local conditions or disperse poorly
[71][72][73][4,120,121]. Epigenetic regulations, which is the basis of plasticity, give the plants the ability to cope with the variability of the habitat conditions.
The seeds in the soil seed bank are constantly adjusting their dormancy to harmonise germination with climatic space and the season of the year
[74][122]. In response to environmental stimuli, seeds show epigenetic changes that, in turn, result in the expression of dormancy-regulating genes. The team of Liu et al.
[75][123] indicated that factors related to the PAF1C (RNA Polymerase II Associated Factor 1) complex, such as VIP4 (Vernalization Independence 4), VIP5, ELF7 (Early Flowering 7), ELF8, HUB1 (Histone Monoubiquitination 1), or RDO2 (Reduced Dormancy 2) are involved in the regulation of the dormancy in seeds and in early flowering. In addition, VIP4, VIP5, ELF7, and ELF8 are required for the expression of
FLC (Flowering locus C), which can be regarded as a seed memory candidate gene due to its association with both flowering and seed dormancy. In turn, the HDA6 and HDA19 histone deacetylases are responsible for the regulation of germination by inhibiting the embryo-specific genes
LEC1 (leafy cotyledon1),
FUS3 (FUSCA3), and
ABI3 (abscisic acid insensitive 3)
[76][77][124,125]. Moreover, had2 and HD2A deacetylases correlate with the expression of the
ELO3 (elongata 3) gene, which encodes a histone acetyltransferase in
Arabidopsis, and the associated
DOG1 (delay of germination 1) gene
[74][122]. HUB1 is a conserved ubiquitin-like protein and is required for the monoubiquitination of histone H2B at lysine 143 (H2BK143)
[78][79][126,127]. It is a prerequisite for the trimethylation of lysine 4 (H3K4me3) and 79 (H3K79me3), which is related to gene activation
[80][128]. Histone H2B monoubiquitination facilitates both transcription elongation and nucleosome refolding, and its loss leads to a reduction in the level of
DOG1 transcripts in seeds, thus contributing to the subsiding of seed dormancy
[81][82][129,130].
The persistence of seeds well adapted to changes in the ecosystem allows them to disperse over time and avoid the beginning of the germination phase until favourable conditions appear
[83][131]. In the evolutionary context, delaying seed germination (bet-hedging strategy) spreads the risk of reproductive failure, which is especially important in an unpredictable environment where the risk of dying before reaching maturity is high
[84][132]. A well-known genetic germination barrier is
DOG1, a key dormancy regulator that determines the optimal temperature for seed germination
[85][133].
DOG1 is specifically expressed in seeds and encodes a protein with unknown molecular functions. It belongs to a small family of proteins in
Arabidopsis containing three conserved domains: PD87616, PD4114, and PD3883
[86][87][134,135]. Nakabayashi et al.
[86][134] showed that DOG1 protein levels in mature seeds correlate with dormancy and remain stable during seed storage.
DOG1 is alternatively spliced to produce four different cDNAs that are combinations of fragments of three exons. The functions of these isoforms remain unknown; however, their relative ratio does not change during seed development
[87][135]. Cyrek et al.
[88][136] showed that, as a result of alternative polyadenylation of the
DOG1 gene, two mRNA variants of this gene are generated, short (sh
DOG1) and long (lg
DOG1). sh
DOG1 in
Arabidopsis is responsible for the production of the DOG1 protein and is, thus, responsible for establishing seed dormancy time
[88][136]. Footit et al.
[74][122] investigated changes in the chromatin of seeds from a soil seed bank and found that both the expression-activating sign H3K4me3 and the repressive sign H3K27me3 play a key role in temporal detection by regulating the expression of the
DOG1 gene. Moreover, modifications of the histone H3 in the form of H3K4me3 and H3K27me3 established near the
DOG1 gene are responsible for the thermal detection mechanism during the dormancy cycle. They found that trimethyl lysine 4 on histone H3 along the
DOG1 gene is stable during dormancy maintenance
[74][122]. The repressive sign H3K27me3 slowly accumulates and accelerates upon exposure to light, ultimately leading to the loss of dormancy. Additionally, Müller et al.
[89][137] focused on the observation of chromatin dynamics in key genes responsible for the regulation of seed dormancy, investigating two opposite signs of histone H3 methylation, i.e., H3K4me3 and H3K27me3. The mutual regulation of these signs was found through the transition from H3K4me3, responsible for the activation of gene expression, to the accumulation of repressive markers in the form of H3K27me3, which, in turn, persisted through the next stage of seedling growth. Thus, the transition to another phase of life is directly reflected in the change in chromatin levels, which is then sustained throughout further development
[89][137].
ABI3 (abscisic acid-insensitive 3), the major regulator of the abscisic acid (ABA) signalling pathway, is a protein transcriptionally regulated at the chromatin level in
Arabidopsis and in yellow cedar seeds. During the transition from dormancy to germination, the chromatin markers change from the active state (H3K4me3) to the repressive state (H3K27me3)
[89][90][91][137,138,139].
DOG1 stimulates temperature-dependent dormancy, thereby influencing the levels of specific miRNAs
[92][140]. Thus,
DOG1 can regulate dormancy by influencing the production and/or function and processing of the miRNAs miR156 and miR172, high levels of which inhibit (miR156) or promote (miR172)
Arabidopsis seed germination at high temperatures
[92][93][140,141]. miR159c is involved in the control of MYB33 and MYB101 transcription factors, which positively regulate the ABA (abscisic acid) signalling pathway
[94][95][142,143]. In addition,
DOG1 influences the expression of genes that code for miRNA processing proteins by inducing the transcript of the dicer-like1 (DCL1) enzyme and the hyponastic leaves1 (HYL1) RNA-binding protein and inhibiting the SERRATE (SE) protein
[92][140]. Huo et al.
[92][140] showed that the
DOG1 gene, involved in determining the seasonal germination time, influences also the flowering time of plants. Consequently, it provides a molecular mechanism that coordinates the response of dormant seeds and flowering plants with the environmental conditions.