Plant material
L. perenne GR3320 (2n = 2x = 14), F. arundinacea subsp. arundinacea (2n = 6x = 42), F. gigantea GR11759 (2n = 6x = 42), F. mairei GR610941 (2n = 4x = 28) were obtained as seeds from Leibniz Institute of Plant Genetics and Crop Plant Research (Gatersleben, Germany) gene bank. Seeds of F. pratensis cv. Fure (2n = 2x = 14) were obtained from Dr. Arild Larson (Graminor, Norway), L. perenne cv. Neptun (2n = 4x = 28) and L. multiflorum cv. Kuri1 (2n = 2x = 14), from Dr. Vladimír Černoch (DLF Seeds, Czech Republic). Plants of L. multiflorum cv. Mitos (2n = 4x = 28), F. pratensis cv. Westa (2n = 4x = 28) and F. glaucescens (2n = 4x = 28) were provided by one of us (DK).
Seeds of barley (Hordeum vulgare) cv. Morex, rye (Secale cereale) cv. Dánkowskie Diament, oats (Avena sativa) cv. Atego were obtained from Leibniz Institute of Plant Genetics and Crop Plant Research gene bank. Seeds of Triticum aestivum cv. Chinese Spring were obtained from Prof. Takashi R. Endo (Kyoto University, Japan) and seeds of Aegilops tauschii were provided by Dr. Valárik (Institute of Experimental Botany, Czech Republic). Seeds of pea (Pisum sativum cv. Ctirad) and rye (Secale cereale cv. Dankovske), which served as internal reference standards in flow cytometric analysis, were provided by one of us (JD).
Estimation of nuclear genome size
Nuclear DNA amounts were determined according to Doležel et al. [65] following the two-step procedure of Otto [66] with modifications. Samples of isolated nuclei stained by propidium iodide were analysed using Sysmex CyFlow Space flow cytometer (Sysmex Partec GmbH, Münster, Germany) equipped with a 532-nm laser. Two reference standards were used to estimate DNA amounts in absolute units. Pea (Pisum sativum cv. Ctirad; 2C = 9.09 pg DNA, [40] served as an internal standard for DNA content estimation in all accessions with the exception of F. mairei, for which rye (Secale cereale cv. Dankovske; 2C = 16.19 pg DNA, [40] was used. Three plants were measured per accession and each plant was analysed three times on three different days. At least 5000 nuclei per sample were analysed. Nuclear DNA content was then calculated from individual measurements following the formula:
2C nuclear DNA content [pg] = 2C nuclear DNA content of reference standard × sample G1 peak mean / standard G1 peak mean
Mean nuclear DNA content (2C) was calculated for each plant. Genome size (1C value) was then determined considering 1 pg DNA equal to 0.978×109 bp [67]. Statistical significance of differences between monoploid (1Cx) genome sizes were determined using one-way ANOVA. The analysis was performed using NCSS 97 statistical software (Statistical Solutions Ltd., Cork, Ireland). The significance level α = 0.01 was used.
Phylogenetic analysis
Phylogenetic analysis of Loliinae subtribe was done based on data published by Catalán et al. [3]. Sequence of ITS regions were downloaded from the NCBI GenBank (GB codes: AF303401-407, AF303410-416, AF303418-419, AF303421-425, AF303428, AF478475-476, AF478478-491, AF478493, AF478498-499, AF519975-981, AF519983, AF532937, AF532939-948, AF532951-952, AF532954, AF532956-960, AF532962-963, AF543514, AF548028, AJ240143, AJ240146, AJ240148, AJ240153, AJ240155-157, AJ240160, AJ240162, AY099007, AY118087-088, AY118090-092, AY118094-096, AY228161). Brachypodium distachyon (GB code AF303339) was used as an outgroup species. The sequences were aligned by MAFFT program v7.029 (--localpair --maxiterate 1000) [68] and phylograms were constructed by PhyML 3.0 [69] implemented in SeaView v5.0.2 [70]. Approximate likelihood ratio test [71] was performed to assess the branch support. Phylogenetic trees were drawn and edited using FigTree program (http://tree.bio.ed.ac.uk/software/figtree/).
Illumina sequencing and data analysis
Genomic DNA was isolated using the NucleoSpin PlantII kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany) following the manufacturer's recommendations and used for preparation of Illumina libraries using Nextera® DNA Sample Preparation Kit (Illumina, San Diego, USA). 50 ng of DNA was fragmented, purified and amplified according to the protocol. DNA concentration in individual libraries was measured using a Qubit fluorometer, adjusted to an equal molar concentration and pooled prior to sequencing. DNA sequencing was done with an Illumina MiSeq using either single or paired end sequencing to produce up to 500 base pair reads. Sequences reads were deposited in the Sequence Read Archive (BioProject ID: PRJNA601325, accessions SAMN13866227, SAMN13866228, SAMN13866229, SAMN13866230, SAMN13866231, SAMN13866232, SAMN13866233, SAMN13866234, SAMN13866235, SAMN13866236).
Illumina reads were trimmed for adapters and for quality using FASTX-toolkit [-q 20 -p 90] (http://hannonlab.cshl.edu/fastx_toolkit/index.html). Detailed characterization of repeat families was performed using stand-alone version of RepeatExplorer pipeline [36] running on IBM server with 16 processors, 100Gb of RAM and 17Tb of disk space. In the first step, comparative analysis of repetitive parts of the genomes was performed using the RepeatExplorer pipeline according to Novák et al. [48]. Random data sets represented the same amount of reads 0.5× coverage of individual accessions and used to reconstruct repetitive elements using graph-based method according Novák et al. [47]. The assembled sequences within each individual cluster were characterized based on the homology searches and other tools useful for repeat characterization (e.g. BLASTN and BLASTX programs, phylogenetic analysis). Tandem organized repeats were identified using Dotter [72].
In the second step, RepeatExplorer pipeline was applied on a merged dataset containing all species marked by specific prefixes to perform comparative analysis [48]. The results of the clustering were then used to create repetitive databases. Databases of Illumina reads and assembled contigs from different types of repetitive DNA elements are publicly available on web site https://olomouc.ueb.cas.cz/en/content/dna-repeats.
Southern hybridization
Genomic DNA corresponding to 3 × 106 copies of a monoploid (1Cx) nuclear genome was digested by HaeIII enzyme (New England Biolabs, Ipswick, Massachusetts, USA). DNA fragments were size-fractionated by electrophoresis in 1.2 % agarose gel and then transferred onto HybondTM N+ nylon membranes (GE Healthcare, Chicago, Illinois, USA). Probes were prepared using F. pratensis genomic DNA as template and PCR with biotin-labelled dUTP (Roche, Mannheim, Germany) and specific primers (Table 3). Southern hybridization was performed at 68°C overnight and hybridization signals were detected by Chemiluminescent Nucleic Acid Module (Thermo Fisher Scientific, Waltham, Massachusetts, USA) according to manufacturer's recommendations with 90 % stringency. Hybridization signals were visualized by chemiluminiscent substrate on Medical X-Ray Film Blue (Agfa HealthCare NV, Mortsel, Belgium).
Droplet digital PCR
Based on the assembled DNA contigs from Fesreba retrotransposon, two restriction endonucleases with unique restriction site in the retrotransposon (HpaI and HpaII) were identified and used for further analysis. 3 µg of genomic DNA was digested according manufacturer’s recommendations (Bio-Rad Laboratories, Hercules, California, USA) and then diluted 1,000-fold to reach starter concentration of 0.06 ng/µl. Droplet Digital PCR experiment was performed using QX200 Droplet Digital PCR machine (Bio-Rad Laboratories) following manufacturer’s recommendations using EvaGreen Supermix (Bio-Rad Laboratories), template DNA and specific primers for Fesreba (Additional file 6: Table S3). Three independent replicates were done for every analyzed accession.
Cytogenetic mapping and immunostaining
Cytogenetic mapping of selected repeats was done by fluorescence in situ hybridization (FISH) on mitotic metaphase plates. Chromosome spreads were prepared according to Křivánková et al. [34] and immunostaining was done according to Neumann et al. [73]. Root tips were collected into ice water for 28 h, washed in LB01 buffer [74], fixed in 3.7% formaldehyde for 25 min and digested using 2 % cellulose, 2 % pectinase and 2 % cytohelicase in 1× PBS for 90 min at 37 °C. After squashing the meristem and coverslip removal, the slides were washed in 1× PBS and then in PBS-Triton buffer (1× PBS, 0.5% Triton X-100, pH 7.4) for 25 min and then again in 1× PBS. For incubation with anti-grass CENH3 primary antibody [75], the slides were washed in PBS-Tween buffer (1× PBS, 0.1 % Tween 20, pH 7.4) for 25 min and then incubated with anti-grass CENH3 primary antibody (diluted 1 : 200 in PBS-Tween) overnight at 4°C. Next day slides were washed 1× PBS, the CENH3 antibody was detected using the anti-Rabbit Alexa Fluor 546 secondary antibody (ThermoFisher Scientific/Invitrogen) diluted 1 : 250 in PBS-Tween buffer, for 1 h at room temperature and washed 1× PBS. Before the FISH procedure,immunofluorescent signals were stabilized using ethanol : acetic acid (3 : 1) fixative and 3.7% formaldehyde for 10 min at room temperature. FISH was performed after three washes in 1 × PBS.
Probes for FISH, derived from RT and LTR regions of Fesreba element, were labelled by digoxigenin-11-dUTP or biotin-16-dUTP (Roche Applied Science) using PCR with specific primers (Table 3). Hybridization mixture consisting of 50% formamide, 10% dextran sulfate in 1× SSC and 1 μg/ml of each labelled probe was added onto slides and denatured at 80°C for 3 min. The hybridization was carried out at 37°C overnight. The sites of probe hybridization were detected using anti-digoxigenin-FITC (Roche Applied Science) and streptavidin-Cy3 (Thermo Fisher Scientific, Waltham, Massachusetts, USA), the chromosomes were counterstained with DAPI and mounted in Vectashield (Vector Laboratories). The slides were examined with Axio Imager.Z2 microscope (Carl Zeiss, Oberkochen, Germany) equipped with Cool Cube 1 (Metasystems, Altlussheim, Germany) camera and appropriate optical filters. The capture of fluorescence signals and merging the layers were performed with ISIS software 5.4.7 (Metasystems) and the final image adjustment was done in Adobe Photoshop 12.0.