Transcriptional activity of transposable elements along an elevational gradient in Arabidopsis arenosa

Abstract

Background: Plant genomes can respond rapidly to environmental changes and transposable elements (TEs) arise as important drivers contributing to genome dynamics. Although some elements were reported to be induced by various abiotic or biotic factors, there is a lack of general understanding on how environment influences the activity and diversity of TEs. Here, we combined common garden experiment with short-read sequencing to investigate genomic abundance and expression of 2245 consensus TE sequences (containing retrotransposons and DNA transposons) in an alpine environment in Arabidopsis arenosa. To disentangle general trends from local differentiation, we leveraged four foothill-alpine population pairs from different mountain regions. Seeds of each of the eight populations were raised under four treatments that differed in temperature and irradiance, two factors varying with elevation. RNA-seq analysis was performed on leaves of young plants to test for the effect of elevation and subsequently of temperature and irradiance on expression of TE sequences.

Results: Genomic abundance of the 2245 consensus TE sequences varied greatly between the mountain regions in line with neutral divergence among the regions, representing distinct genetic lineages of A. arenosa. Accounting for intraspecific variation in abundance, we found consistent transcriptomic response for some TE sequences across the different pairs of foothill-alpine populations suggesting parallelism in TE expression. In particular expression of retrotransposon LTR Copia (e.g. Ivana and Ale clades) and LTR Gypsy (e.g. Athila and CRM clades) but also non-LTR LINE or DNA transposon TIR MuDR consistently varied with elevation of origin. TE sequences responding specifically to temperature and irradiance belonged to the same classes as well as additional TE clades containing potentially stress-responsive elements (e.g. LTR Copia Sire and Tar, LTR Gypsy Reina).

Conclusions: Our study demonstrated that the A. arenosa genome harbours a considerable diversity of TE sequences whose abundance and expression response varies across its native range. Some TE clades may contain transcriptionally active elements responding to a natural environmental gradient. This may further contribute to genetic variation between populations and may ultimately provide new regulatory mechanisms to face environmental challenges.

Keywords: Alpine environment; Arabidopsis arenosa; Common garden experiment; Parallelism; RNA-seq; Transposable elements.

Fig. 1

a Original locations of the eight populations from the four mountain regions used in this study. Dots coloured by ecotype (black = foothill, grey = alpine ecotype). b Principal component analysis based on the two environmental variables, temperature and irradiance, of the original sampling sites, coloured by ecotype (black = foothill, grey = alpine ecotype). We estimated the average values of temperature and irradiance (Photosynthetic Active Radiation [PAR]) over April, May and June that corresponds to the growth period of A. arenosa. The two variables were obtained from the high-resolution climate database SolarGIS, version 1.9, operated by GeoModel Solar (Bratislava, Slovakia). NT = Niedere Tauern (Austria), FG = Făgăraș (Romania), VT = Vysoké Tatry (Slovakia), ZT = Západné Tatry (Slovakia)

Fig. 2

Multidimensional scaling plot showing the level of similarity in (a) genomic abundance (N = 71 individuals) and (b) and (c) in expression (N = 96 individuals) of the 2245 consensus sequences of Arabidopsis arenosa from studied populations. Each symbol represents one individual, in panels (a) and (b) triangles depict alpine and circles foothill ecotype and are coloured by region and in panel (c) symbols are coloured by treatment

Fig. 3

Number of consensus TE sequences differentially expressed between the foothill and alpine ecotype for each region and their overlap. Significance of each intersection tested by Fischer’s exact test is indicated by ** P < 0.01, * P < 0.05, ⁽*⁾ P < 0.1 and by different colour bars