Environmental and genealogical effects on DNA methylation in a widespread apomictic dandelion lineage.

DNA methylation in plant genomes occurs in different sequences and genomic contexts that have very different properties. DNA methylation that occurs in CG (mCG) sequence context shows transgenerational stability and high epimutation rate, and can thus provide genealogical information at short time scales. However, due to meta-stability and because mCG variants may arise due to other factors than epimutation, such as environmental stress exposure, it is not clear how well mCG captures genealogical information at micro-evolutionary time scales. Here, we analysed DNA methylation variation between accessions from a geographically widespread, apomictic common dandelion (Taraxacum officinale) lineage when grown experimentally under different light conditions. Using a reduced-representation bisulphite sequencing approach, we show that the light treatment induced differentially methylated cytosines (DMCs) in all sequence contexts, with a bias towards transposable elements. Accession differences were associated mainly with DMCs in CG context. Hierarchical clustering of samples based on total mCG profiles revealed a perfect clustering of samples by accession identity, irrespective of light conditions. Using microsatellite information as a benchmark of genetic divergence within the clonal lineage, we show that genetic divergence between accessions correlates strongly with overall mCG profiles. However, our results suggest that environmental effects that do occur in CG context may produce a heritable signal that partly dilutes the genealogical signal. Our study shows that methylation information in plants can be used to reconstruct micro-evolutionary genealogy, providing a useful tool in systems that lack genetic variation such as clonal and vegetatively propagated plants.

DNA methylation in clonal duckweed (Lemna minor L.) lineages reflects current and historical environmental exposures.

Environmentally induced DNA methylation variants may mediate gene expression responses to environmental changes. If such induced variants are transgenerationally stable, there is potential for expression responses to persist over multiple generations. Our current knowledge in plants, however, is almost exclusively based on studies conducted in sexually reproducing species where the majority of DNA methylation changes are subject to resetting in germlines, limiting the potential for transgenerational epigenetics stress memory. Asexual reproduction circumvents germlines, and may therefore be more conducive to long-term inheritance of epigenetic marks. Taking advantage of the rapid clonal reproduction of the common duckweed Lemna minor, we hypothesize that long-term, transgenerational stress memory from exposure to high temperature can be detected in DNA methylation profiles. Using a reduced representation bisulphite sequencing approach (epiGBS), we show that temperature stress induces DNA hypermethylation at many CG and CHG cytosine contexts but not CHH. Additionally, differential methylation in CHG context that was observed was still detected in a subset of cytosines, even after 3-12 generations of culturing in a common environment. This demonstrates a memory effect of stress reflected in the methylome and that persists over multiple clonal generations. Structural annotation revealed that this memory effect in CHG methylation was enriched in transposable elements. The observed epigenetic stress memory is probably caused by stable transgenerational persistence of temperature-induced DNA methylation variants across clonal generations. To the extent that such epigenetic memory has functional consequences for gene expression and phenotypes, this result suggests potential for long-term modulation of stress responses in asexual plants.

epiGBS2: Improvements and evaluation of highly multiplexed, epiGBS-based reduced representation bisulfite sequencing.

Several reduced-representation bisulfite sequencing methods have been developed in recent years to determine cytosine methylation de novo in nonmodel species. Here, we present epiGBS2, a laboratory protocol based on epiGBS with a revised and user-friendly bioinformatics pipeline for a wide range of species with or without a reference genome. epiGBS2 is cost- and time-efficient and the computational workflow is designed in a user-friendly and reproducible manner. The library protocol allows a flexible choice of restriction enzymes and a double digest. The bioinformatics pipeline was integrated in the Snakemake workflow management system, which makes the pipeline easy to execute and modular, and parameter settings for important computational steps flexible. We implemented bismark for alignment and methylation analysis and we preprocessed alignment files by double masking to enable single nucleotide polymorphism calling with Freebayes (epiFreebayes). The performance of several critical steps in epiGBS2 was evaluated against baseline data sets from Arabidopsis thaliana and great tit (Parus major), which confirmed its overall good performance. We provide a detailed description of the laboratory protocol and an extensive manual of the bioinformatics pipeline, which is publicly accessible on github (https://github.com/nioo-knaw/epiGBS2) and zenodo (https://doi.org/10.5281/zenodo.4764652).

The methylome of Biomphalaria glabrata and other mollusks: enduring modification of epigenetic landscape and phenotypic traits by a new DNA methylation inhibitor.

BACKGROUND: 5-Methylcytosine (5mC) is an important epigenetic mark in eukaryotes. Little information about its role exists for invertebrates. To investigate the contribution of 5mC to phenotypic variation in invertebrates, alteration of methylation patterns needs to be produced. Here, we apply new non-nucleoside DNA methyltransferase inhibitors (DNMTi) to introduce aleatory changes into the methylome of mollusk species. RESULTS: Flavanone inhibitor Flv1 was efficient in reducing 5mC in the freshwater snails Biomphalaria glabrata and Physa acuta, and to a lesser degree, probably due to lower stability in sea water, in the oyster Crassostrea gigas. Flv1 has no toxic effects and significantly decreased the 5mC level in the treated B. glabrata and in its offspring. Drug treatment triggers significant variation in the shell height in both generations. A reduced representation bisulfite-sequencing method called epiGBS corroborates hypomethylation effect of Flv1 in both B. glabrata generations and identifies seven Differential Methylated Regions (DMR) out of 32 found both in Flv1-exposed snails and its progeny, from which 5 were hypomethylated, demonstrating a multigenerational effect. By targeted bisulfite sequencing, we confirmed hypomethylation in a locus and show that it is associated with reduced gene expression. CONCLUSIONS: Flv1 is a new and efficient DNMTi that can be used to induce transient and heritable modifications of the epigenetic landscape and phenotypic traits in mollusks, a phylum of the invertebrates in which epigenetics is understudied.

Genetic fine-mapping of DIPLOSPOROUS in Taraxacum (dandelion; Asteraceae) indicates a duplicated DIP-gene.

BACKGROUND: DIPLOSPOROUS (DIP) is the locus for diplospory in Taraxacum, associated to unreduced female gamete formation in apomicts. Apomicts reproduce clonally through seeds, including apomeiosis, parthenogenesis, and autonomous or pseudogamous endosperm formation. In Taraxacum, diplospory results in first division restitution (FDR) nuclei, and inherits as a dominant, monogenic trait, independent from the other apomixis elements. A preliminary genetic linkage map indicated that the DIP-locus lacks suppression of recombination, which is unique among all other map-based cloning efforts of apomeiosis to date. FDR as well as apomixis as a whole are of interest in plant breeding, allowing for polyploidization and fixation of hybrid vigor, respectively. No dominant FDR or apomixis genes have yet been isolated. Here, we zoom-in to the DIP-locus by largely extending our initial mapping population, and by analyzing (local) suppression of recombination and allele sequence divergence (ASD). RESULTS: We identified 24 recombinants between two most closely linked molecular markers to DIP in an F1-population of 2227 plants that segregates for diplospory and lacks parthenogenesis. Both markers segregated c. 1:1 in the entire population, indicating a 1:1 segregation rate of diplospory. Fine-mapping showed three amplified fragment length polymorphisms (AFLPs) closest to DIP at 0.2 cM at one flank and a single AFLP at 0.4 cM at the other flank. Our data lacked strong evidence for ASD at marker regions close to DIP. An unexpected bias towards diplosporous plants among the recombinants (20 out of 24) was found. One third of these diplosporous recombinants showed incomplete penetrance of 50-85% diplospory. CONCLUSIONS: Our data give interesting new insights into the structure of the diplospory locus in Taraxacum. We postulate a locus with a minimum of two DIP-genes and possibly including one or two enhancers or cis-regulatory elements on the basis of the bias towards diplosporous recombinants and incomplete penetrance of diplospory in some of them. We define the DIP-locus to 0.6 cM, which is estimated to cover approximately 200-300 Kb, with the closest marker at 0.2 cM. Our results confirm the minor role of suppression of recombination and ASD around DIP, making it an excellent candidate to isolate via a chromosome-walking approach.