Predicting heterosis in F1 hybrids from parental DNA methylomes
SFB924 project, cooperation with F. Johannes, TU München
A pilot study using 19 different epigenetic hybrid (epiHybrid) populations revealed widespread heterosis in the epiHybrids (Lauss et al. (2018) Plant Physiol. https:// doi.org/10.1104/pp.17.01054), indicating that methylation differences between the parental lines are sufficient to trigger heterosis in hybrid offspring, independently of DNA sequence determinants. One putative parental DMR is associated with heterosis in plant height. We expect that more DMR-trait association will be detectable with larger samples sizes. To this end we created 382 diverse epiHybrid populations by crossing a male-sterile Col-0 wildtype (Col-wt) as maternal parent to 382 epiRILs as the paternal parents.
In the first phenotyping phase, we conducted three cultivation experiments using the 382 different epiRIL hybrids in parallel with their corresponding 382 paternal epiRILs and recurrent maternal Col-wt line. In each of the three cultivation experiments 6 individuals per line were grown and phenotyped (4608 individuals per cultivation experiment; 13824 individuals in total across the three cultivation experiments). Plants were phenotyped for growth and early developmental traits (i.e. architectural traits such as leaf area, rosette diameter, convex hull of rosette, rosette compactness, etc., as well as color- and fluorescence-related traits). Based on our evaluation of these early developmental phenotypes, we selected a set of 192 epiRILs and their parental lines for a second phenotyping phase. In this second phase, we will conduct three cultivation experiments using 192 different epiRIL hybrids in parallel with their corresponding 192 paternal epiRILs and recurrent maternal Col-wt line. In each of the three cultivation experiments, six individuals per line will be grown and phenotyped (2304 individuals per cultivation experiment; 6912 individuals in total across the three cultivation experiments). Plants will be grown on 6 well trays until 27 days after sowing and phenotyped for flowering time (bolting) in addition to assessing, again, the early vegetative growth and developmental traits. Plant material from the 192 trios (i.e. 192 epiHybrids + 192 epiRILs + 1 Col-wt) will be used subsequently for high-throughput molecular analyses of DNA methylation and gene expression (transcriptome).
An important by-product of this project is the provision of high-resolution methylome sequences of 192 epiRILs. The provision of cytosine resolution methylomes will be an important tool for future plant epigenetics studies using this experimental system. The additional methylome measurements obtained will be used to re-calibrate the existing DMR linkage map (Colomé-Tatché et al. (2012) PNAS 198:16240-16245). This should improve mapping resolution, and facilitate deeper insights into the epigenetic architecture underlying complex traits.
The re-calibrated linkage map will be employed to identify heritable DMRs between the selected epiRILs and the Col-wt parental line. Second, these DMRs will be used to perform an epigenome-wide linkage scan to search for parental genomic regions that are predictive of heterosis in the epiHybrids. We will perform this analysis for all phenotypes that show heterosis in a large number of independent epiHybrid populations (as identified from 3.4.3), as this is indicative of these effects being caused by common epialleles that segregate among the epiRIL parental lines.