Warmer temperature during asexual reproduction induce methylome, transcriptomic, and lasting phenotypic changes in Fragaria vesca ecotypes

Abstract

Plants must adapt with increasing speed to global warming to maintain their fitness. One rapid adaptation mechanism is epigenetic memory, which may provide organisms sufficient time to adapt to climate change. We studied how the perennial Fragaria vesca adapted to warmer temperatures (28°C vs. 18°C) over three asexual generations. Differences in flowering time, stolon number, and petiole length were induced by warmer temperature in one or more ecotypes after three asexual generations and persisted in a common garden environment. Induced methylome changes differed between the four ecotypes from Norway, Iceland, Italy, and Spain, but shared methylome responses were also identified. Most differentially methylated regions (DMRs) occurred in the CHG context, and most CHG and CHH DMRs were hypermethylated at the warmer temperature. In eight CHG DMR peaks, a highly similar methylation pattern could be observed between ecotypes. On average, 13% of the differentially methylated genes between ecotypes also showed a temperature-induced change in gene expression. We observed ecotype-specific methylation and expression patterns for genes related to gibberellin metabolism, flowering time, and epigenetic mechanisms. Furthermore, we observed a negative correlation with gene expression when repetitive elements were found near (±2 kb) or inside genes. In conclusion, lasting phenotypic changes indicative of an epigenetic memory were induced by warmer temperature and were accompanied by changes in DNA methylation patterns. Both shared methylation patterns and transcriptome differences between F. vesca accessions were observed, indicating that DNA methylation may be involved in both general and ecotype-specific phenotypic variation.

Figure 1

Experimental setup and phenotypic responses under common garden-conditions in Fragaria vesca plants after propagation at different temperatures for up to three asexual generations. (A) Experimental setup: plants were propagated for three asexual generations (AS1–3) through stolon formation at 18°C (blue boxes) or 28°C (red boxes). Leaf samples for DNA and RNA sequencing were collected from AS1 plants (at 18°C only) and AS3 plants (at both temperatures). Phenotypic observations of AS1 and AS3 plants were made under common garden-conditions following short-day (SD) treatment (green boxes). (B-D) Total number of stolons (B), flowering time (C), and petiole length (D) in the ES12, ICE2, IT4, and NOR2 ecotypes. Plants were propagated at 18°C or 28°C and phenotypes were scored under common garden-conditions after flower-inducing short-day (SD) treatment. Flowering time was measured as days from the start of SD treatment until the first flower had opened completely. Flowering time could not be determined in ES12 plants since these could not be induced to flower. Arrowhead in the upper image indicates stolon; the bracket in the lower image indicates petiole length. Brackets and asterisks indicate significant differences from Wilcoxon tests: ^* 0.01 ≤ p < 0.05; ^** 0.001 ≤ p < 0.01; ^*** 0.0001 ≤ p < 0.001. § p = 0.06. Error bars show 95% confidence intervals. Box plots also show median values (n = 10) and the interquartile range (difference between the 75th and 25th percentiles).

Figure 2

DNA methylation landscape in leaves of four Fragaria vesca ecotypes propagated for three asexual generations at 18°C or 28°C. Ecotypes are from Spain (ES12), Iceland (ICE2), Italy (IT4), and Norway (NOR2). All methylation data show the CGN, CHG, and CHH methylation contexts (where N can be any base and H can be any base other than G). Three biological replicates were used for each ecotype and temperature combination. (A) Principal component analysis of methylation in each ecotype grown at 18°C (blue symbols) and 28°C (red symbols). Black crosses show projections of the data points on the xy-plane. Red lines indicate the x, y and z axes. (B) Methylation patterns along all seven F. vesca chromosomes for all ecotypes and both temperatures, using a 50 kb window. Lanes 1–8 show methylation levels for each ecotype-temperature combination. Connecting lines in the middle indicate coding regions (CDS) regions that showed synteny. (C-E) Methylation level along protein coding genes (C), REs (D), and pseudogenes (E) in the NOR2 ecotype. Each genomic feature and regions 2 kb up- and downstream to these were divided into 20 pieces to calculate methylation levels (methylated reads/total reads). Colored lines show different combinations of temperature and methylation contexts.

Figure 3

Chromosomal and genic distribution patterns of DMRs in Fragaria vesca leaves propagated for three asexual generations at 18°C or 28°C. (A) Pattern of DMRs on each F. vesca (Fv) chromosome in the NOR2 ecotype. Lanes 1–7 show gene density, DMR density, and methylation level (%; methylated reads/total reads, using a 50 kb window) for different methylation contexts (CGN, CHG, and CHH, where N can be any base and H can be any base other than G). Connecting lines in the middle indicate coding regions (CDS) regions that showed synteny. (B) Distribution of DMRs for different methylation contexts across different genomic features in four F. vesca ecotypes (ES12, ICE2, NOR2, and IT4). (C) Distribution of DMRs for different methylation contexts according to their distance (kilobase-pairs in the 5′-3′ direction) from transcription start sites (TSS) in different ecotypes. Three biological replicates were used for each ecotype and temperature combination.

Figure 4

DMRs and genomic features inside DMR peaks in leaves of different Fragaria vesca ecotypes propagated for three asexual generations at 18°C or 28°C. Distribution of DMRs, pseudogenes, REs, and genes along eight 50 kb wide DMR peaks for the CHG methylation context (where H can be any base other than G) in four F. vesca (Fv) chromosomes. The x-axis indicates location along the DMR peak. Red and blue bars represent hyper- and hypo-methylated DMRs, respectively. Coloured arrows indicate pseudogenes (yellow), REs (purple), and genes (brown). Asterisks indicate predicted TARGET OF PROTEIN FOR XKLP2 (TPX2) genes.

Figure 5

Relationship between DNA methylation and gene expression in Fragaria vesca ecotypes propagated for three asexual generations at 18°C or 28°C. (A) Venn diagrams showing numbers of differentially methylated genes (DMGs) for different ecotypes (ES12, ICE2, NOR2, IT4) and methylation contexts (CGN, CHG, CHH, where N can be any base and H can be any base other than G). (B) Volcano plot showing differentially expressed genes (DEGs) in NOR2 plants propagated at 28°C vs. 18°C. Blue and red dots show down- and upregulated genes, respectively. Dashed lines delineate P-value ≤0.05 and |FoldChange | ≥ 1.5 calculated from three biological replicates. (C) Venn diagrams showing overlap between DEGs and DMGs in the different ecotypes. (D) Enriched GO terms among DEDMGs in the different ecotypes. The x-axis indicates the number of DEDMGs per GO term. Color shading indicates fold enrichment at 28 vs. 18°C. (E) Venn diagrams showing numbers of DEDMGs in all ecotypes.

Figure 6

Functional analysis of DEDMGs in Fragaria vesca ecotypes propagated for three asexual generations at 18°C or 28°C. Graphs show expression and methylation status of four DEDMGs on F. vesca chromosome 4 (FvH4) that are shared by all four ecotypes (ES12, ICE2, NOR2, IT4). Red and blue bars represent hyper- and hypomethylated DMRs, respectively, in different methylation contexts (CHG, CGN, CHH, where N can be any base and H can be any base other than G). ^*, +, and # indicate presence of CGN, CHG, and CHH DMRs, respectively, at the positions indicated by the black vertical lines.