Portal de Periódicos da CAPES

The diversifying field of plant epigenetics

2018; Wiley; Volume: 217; Issue: 3 Linguagem: Inglês

10.1111/nph.14985

1469-8137

Katrin Heer, Jeannie Mounger, M. Teresa Boquete, Christina L. Richards, Lars Opgenoorth,

Plant Genetic and Mutation Studies

Plant epigenetics has become a fast moving research field in the plant sciences. It gained impetus in recent years thanks to developments in plant genetics and sequencing technologies. Two relatively distinct lines of research have been distinguished in the recent literature (Richards et al., 2017). On the one hand, studies conducted on a few model organisms have explored molecular mechanisms using the finest of novel sequencing technologies under the controlled conditions of high-tech laboratories. On the other hand, studies interested in the ecological and evolutionary relevance of epigenetics in real-world settings take on the challenges of working with nonmodel species in complex environments, but have been limited by less powerful genetic analysis tools. Both lines of research now depend heavily on the application of experimental and bioinformatic tools to unravel genetic and epigenetic effects, and to decipher novel sequencing data. In Vienna, researchers from across the fields of ecology, molecular biology, and bioinformatics came together during the 40th New Phytologist Symposium to present the latest areas of research in plant ecological epigenetics. The work presented spanned the breadth of molecular model and nonmodel species as well as ecologically interesting settings. The symposium consisted of 37 oral talks and 39 posters, which altogether related to research on 10 annual plant species, 19 perennial plant species, and nine tree species (Fig. 1). Many studies in plant epigenetics are motivated by understanding the dynamics of epigenetic variation, focusing largely on DNA methylation. The keynote speaker, Frank Johannes (Technical University of Munich, Germany), opened the symposium with a presentation of methylation divergence over time in mutation accumulation (MA) lines of Arabidopsis thaliana. He discovered that the amount of divergence in methylation patterns depends on genomic features, with higher epimutation rates in genes than in transposable elements (TEs). The distinct methylation patterns that resulted from spontaneous epimutations in MA lines were strongly correlated with diversity patterns found in natural accessions. Using methylation site frequency spectrum (mSFS) models, he derived that epimutational processes shape gene-body methylation diversity among natural accessions over evolutionary time-scales. Bob Schmitz (University of Georgia, USA) argued that although epimutations are common, larger regions that consist of multiple consecutive epimutations (differentially methylated cytosines) are rare. His laboratory determined that 99.99% of methylated regions were faithfully inherited over selfing generations of A. thaliana, which permits the reconstruction of epigenotype maps (see Hofmeister et al., 2017). Above the species level, Danelle Seymour (University of California, USA) also found that methylation patterns differ between gene bodies and TEs across three Brassicaceae and across eight grass species. However, she determined that much of the variation of methylation found in TEs was not conserved among species. Conversely, methylation in gene bodies was highly conserved and associated with patterns of gene expression. At a genome-wide level, Conchita Alonso (Estación Biológica de Doñana, CSIC, Spain) presented global methylation differences among species, and found a strong phylogenetic signal where significant portions of the variance were explained by taxonomic order and family. While several of the speakers presented evidence that a large portion of DNA methylation seems to be determined by genomic variation, this body of work did not exclude the possibility of epigenetic variation that is important for response to environmental factors. A variety of research projects in controlled stress experiments and in natural settings explored these possibilities. A number of experiments have been conducted in recent years that manipulate environmental conditions, or apply chemical compounds while controlling for additional environmental factors. Often, these experiments have been carried out in systems with limited genetic variability to specifically test for the effects of a given abiotic stress on epigenetic variation and its inheritance in subsequent generations. Several studies have found that environmental treatments applied in relatively short-term experiments do not result in long-term changes of methylation patterns inherited across generations. In studies on apomictic dandelions, Koen Verhoeven (Netherlands Institute of Ecology, the Netherlands) discovered that the changes in DNA methylation that were induced by the application of jasmonic or salicylic acid were most strongly inherited in the first offspring generation, but weakened in subsequent generations, with strong effects limited to very few loci. Similarly, Jose Gutierrez Marcos (University of Warwick, UK) found that transgenerational effects of salt stress in A. thaliana were already lost in the second generation without the salt treatment. His team revealed that the differentially methylated regions (DMRs) that appeared to be induced by hyperosmotic stress, overlapped with the DMRs that had been repeatedly found in MA lines, indicating the existence of 'epigenetically labile' regions in the genome rather than directed changes in loci that are associated with response to salt stress (Wibowo et al., 2016). Similarly, Steve Eichten (Australian National University, Australia) detected no conserved DMRs for response to cold, UV or heat stress in maize (Eichten & Springer, 2015), but suggested that an increased stochasticity in methylation under stress, rather than directed changes, might be important for the stress response in plants. This increase in DNA methylation variation might contribute to a general increase of phenotypic variation that selection can act on (Richards et al., 2012). However, Stephane Maury (University of Orleans, France) investigated the response to drought stress in shoot apical meristems of poplar clones, and learnt that a large portion of the genes that were down-regulated and hypomethylated during re-watering were associated with phytohormones. Hypomethylated lines were found to be more resistant towards cavitation under severe drought. Further, this hypomethylation response was stable over long time periods in poplar trees, supporting the notion that epigenetic changes contribute to resilience to long-term drought stress. Vítek Latzel (Institute of Botany of the Czech Academy of Sciences, Czech Republic) presented a drought study in the clonal herbaceous plant Trifolium repens, indicating that varying levels and time periods of drought might determine whether epigenetic variation contributes to the stress response (González et al., 2017). Together, these studies suggest that short-term stress under controlled conditions might not result in the same epigenetic signatures as long-term exposure to natural environments. In addition to abiotic conditions, there is growing evidence that biotic interactions may also have important epigenetic components. Isabelle Fudal (BIOGER, France) demonstrated chromatin-based control of effector genes in fungal genomes that infect rapeseed. Claude Becker (Gregor Mendel Institute, Austria) showed that allelochemicals generated by grasses act as inhibitors of histone deacetylases, which resulted in hyper-acetylated regions (Venturelli et al., 2015). His examination of tolerance towards these allelochemicals in natural accessions of A. thaliana may provide a more comprehensive picture of the importance of this mechanism in the wild. Working in Acacia trees, Rebecca Kartzinel (Brown University, USA) presented initial insights into a potential role of epigenetic mechanisms for the plastic changes that are required to develop extrafloral nectaries and domatia for mutualistic ants. These studies exemplify that epigenetic mechanisms might be relevant for a variety of diverse interactions of plants with other plants, animals or fungi that call for further exploration. Given the lack of resources for nonmodel plants, it has been difficult to isolate the importance of epigenetic modifications under natural conditions. One successful approach has been to take advantage of the resources available in model species and run experiments on them under field conditions. For example, Jo Hepworth (John Innes Centre, UK) investigated the regulation of vernalization and flowering in A. thaliana in a common garden exposed to natural temperature fluctuations. She found that low and high temperatures over the course of the day explain variation in the expression levels of FLOWERING LOCUS C (FLC) and VERNALIZATION INSENSITIVE 3 (VIN3). Thierry Halter (Institut de Biologie de l′Ecole Normale Superieure, France) built on previous findings that certain DMRs of ros1 mutants of A. thaliana resulted in increased resistance to Pseudomonas syringae infections (Dowen et al., 2012). Halter discovered that wild accessions from high altitudes had similar methylation patterns as the mutant in the promoter regions of ROS1-SENSITIVE ELICITED (RSE) genes (e.g. RESISTANCE METHYLATED GENE 1 (RMG1), and the zinc cluster protein RSE2). Since P. syringae infections increase the probability of frost damage, this could be an important mechanism for freezing tolerance in this and other taxa. Manu Dubin (Oxford Software Ltd, New Zealand) presented studies on both A. thaliana and barley where CG methylation in gene bodies was positively correlated with the latitude of origin. In A. thaliana accessions that originated from colder environments, genes with higher levels of gene body methylation were more highly expressed possibly due to a general change in transcriptional processes (Dubin et al., 2015). In addition to studies in A. thaliana, a number of studies took advantage of study systems with reduced genetic variation in order to isolate changes in DNA methylation. In clonal alligator weed in China, Yupeng Geng (Yunnan University, China) found that methylation patterns were strongly clustered according to the sampling site even from collections in different years. However, after cultivation in a common environment, these differences were reduced over time, and methylation patterns converged. In a related study, Steve Eichten focused on Brachypodium distachyon families with minimal genetic variation that are widely distributed across Turkey, and thus, exposed to diverse environmental conditions. He found significant variation in flowering time after cultivating seedlings in growth chambers and aims to test how much of this variation can be explained by differences in methylation patterns within families. Many insights in plant epigenetics have been driven by the development of study systems, and by technological and bioinformatic advances in recent years. Future research in plant epigenetics will critically depend on expanding these resources beyond the few model species that currently dominate the research field (Richards et al., 2017). Many studies presented during the symposium took advantage of study systems with low or almost no genetic diversity. For example, A. thaliana epigenetic recombinant inbred lines (epiRILs) were used to determine whether differences based on methylation variation translates to stress responses that could be selected for (Vítek Latzel), or result in differences in resistance towards the oomycete parasite Hyaloperonospora arabidopsidis (Ritushree Jain, University of Sheffield, UK). A number of studies used nonmodel plants that reproduce clonally or are apomictic, such as Taraxacum, Brachypodium, Alternanthera and Trifolium. In addition, approaches for artificially reducing methylation have been refined recently to facilitate future studies. For example, Vítek Latzel found that repeated spraying of established plants with azacytidine reduced methylation in a way that is similar to germinating seeds directly in the chemical, but with fewer side effects. Etienne Bucher (INRA Anger, France) developed an approach to induce a targeted activation of stress-responsive TEs (called jumping TEs) with a combination of zebularine and alpha-amanitin treatment (Thieme et al., 2017). This approach greatly reduced methylation, and led to an activation of TEs via the inhibition of DNA polymerase II (Pol II). Using this technique, Bucher created lines of A. thaliana and maize with a high number of novel insertions of ONSEN, a heat responsive copia-type retrotransposon. These lines were characterized by high phenotypic variation. The mobilization of TEs under stress might not only result in novel breeding strategies, but might also have implications for fast and directed evolutionary processes that should be further explored in the future. Bob Schmitz (University of Georgia, USA) introduced another approach where the expression of a human ten-eleven translocation (TET) enzyme was used to generate epialleles through random demethylation in A. thaliana (Ji et al., 2017). Novel sequencing technologies and decreasing costs have opened many opportunities for nonmodel species to move from anonymous methylation-sensitive amplified fragment length polymorphisms (MS-AFLP or MSAP), towards reduced representation bisulphite sequencing (RRBS), targeted bisulphite sequencing or even whole genome bisulphite sequencing (WGBS) when genomic resources are available (Richards et al., 2017). Already, methylomes are available for about 50 angiosperm species. However, as Conchita Alonso pointed out, these resources are not evenly distributed across the diversity of all plant species. Eriko Sasaki (Gregor Mendel Institute, Austria) presented the resources for A. thaliana, where WGBS data are now available for more than 1000 accessions from the 1001 Epigenome project (Kawakatsu et al., 2016). Christoph Bock (CeMM Research Centre for Molecular Medicine, Vienna, Austria) presented single cell bisulphite sequencing, and the reconstruction of epigenetic landscapes (Farlik et al., 2016), which are currently applied in medical research, and may be adopted in plant sciences in the near future. Still, Bock stressed that even in a model species such as humans, RRBS in large-scale studies is a useful first step. Finally, bioinformatic tools that facilitate the analysis of bisulphite sequencing data and the detection of DMRs are urgently needed (Richards et al., 2017). Peter Stadler (University of Leipzig, Germany) presented his DMR caller metilene that relies on using a simple score and a fast segmentation algorithm (Jühling et al., 2016). The approach presented by Jörg Hagmann (Computomics GmbH, Germany) applies a Hidden Markov Model (HMM) that first identifies methylated regions (Hagmann et al., 2015). Maria Colomé Tatché (Helmholtz Centre Munich, Germany) presented a tool for the imputation of methylation data (METHimpute https://github.com/ataudt/methimpute) that is frequently affected by missing information. METHimpute applies a HMM for the prediction of the methylation status and level across the genome. Finally, Christoph Bock emphasized the importance of complete open access to comprehensive data sets and information on pipelines and code that should accompany publications. An open access policy allows for better reproducibility and comparability of studies. Researchers should also strive to increase the statistical power of plant epigenetic studies, which lag far behind that of medical studies. To date, most plant epigenetic studies in ecological contexts have examined DNA methylation, while studies in model organisms have revealed that histone modifications and small RNAs are also important elements of epigenetic regulation. This indicates that more comprehensive research agendas need to consider not only DNA methylation, but all relevant components of the 'extended genotype' which, according to Steve Eichten includes heritable chromatin factors, behaviour of TEs and insertion and deletion polymorphisms, as well as the effects of 'genomic shock' like polyploidization events. During the discussions it became apparent that delegates disagreed on whether technological advances are closing or enlarging the gap between studies with a more ecological vs more molecular mechanism approach. Many ecological epigeneticists have started to create genomic resources for their study species and integrate novel tools in their systems. However, it remains challenging to keep up with the pace of technological advances in model species. This is especially true since the number of laboratory groups working on a given nonmodel plant species is much smaller than in model plant species (Fig. 1), and ecological epigenetics endeavours to explore responses in complex environments. Molecular biologists at the symposium also acknowledged that the ultimate goal should be to understand the causes and consequences of epigenetics in natural settings. It was clear that both disciplines would greatly benefit from closer collaboration. Thanks to Helen Pinfield-Wells and New Phytologist Central Office, and to Ovidiu Paun and the staff at the University of Vienna and the Botanical Garden for organizing and planning this successful symposium. Thanks also to the New Phytologist Trust for providing travel grants to several graduate student attendees including J.M. Although the authors could not address every meeting presentation in this report, they would like to thank all attendees for their inspiring presentations and discussion during the symposium.