Epigenetics and hidden heritability in tomato

The information content of eukaryotic genomes is based primarily on the DNA sequence but there are additional layers that are dependent on methylation of cytosine residues in the DNA and modification of the histone proteins that bind to DNA in the chromosomes. A primary function of this epigenetic information is to protect against DNA parasites - transposable elements - that colonise genomes and have mutagenic properties.

Epigenetics, therefore, is part of the arms race between the host genomes and these DNA parasites but, like all arms races,the dynamics are not simple. The host may even benefit if a parasite has features or effects that increase the host's fitness and, conversely, the parasite may also benefit if the defense systems of the host reduce the damage caused by parasitism (the parasite is dependent on the well-being of the host).

Evidence of this complex dynamic relationship is from the phenotype of DNA methylation mutants in plants in which DNA methylation is lost throughout the genome. There is not only loss of transposon defense but additionally disrupted growth and development of the plant and sterility. Loss of the DNA methylation in localized parts of the genome, however, may have more specific effects including some that are beneficial.

These observations prompt the hypothesis that epigenetics may account for the hidden heritability that is well known to breeders of crops. Hidden heritability is best illustrated with quantitative traits for which much of the heritable variation between plants cannot be linked to genetic markers. In some instances this hidden heritability or missing inheritance can account for a substantial part of the variation.

Much of the molecular understanding of epigenetics in plants is from the model plant Arabidopsis. Extensive resources that have facilitated experimentation with this plant and it has led to good understanding of the various epigenetic mechanisms. The rapid progress with this model species, however, is partly because the genome is small and has many fewer transposable elements than most other species, including crops.

Perturbation of this reduced epigenome, therefore, has rather mild effects on the growth and development of the plant.

In plants with larger, transposon-rich genomes there are more profound effects on growth and development consistent with the epigenetic transposon defense overlapping with the normal function of the genome. Unfortunately, with the standard set of DNA methylation mutants, the plants may be sterile and be difficult to use experimentally.

To address this problem we have developed an approach in tomato with a DNA methylation mutant that causes limited disruption of the epigenome. In our preparatory work we have isolated and characterized such a mutant in the DNA methylation gene CMT3a. The partial effect allows recovery of fertile mutant plants so that we can implement a detailed analysis of the CMT3a-dependent epigenome and its effects.

In the first part of the proposed project we will fully characterize the CMT3a-dependent epigenome and its influence on gene expression. In our preliminary evidence we show that some regions of the genome lose their DNA methylation in the mutant plans and that they retain this hypomethylated state in backcrossed progeny to wild type plants whereas other regions regain the normal levels of DNA methylation.

Integrating these findings with analyses of gene expression in cmt3a and its progeny will identify features in the genome that correlate with heritability of the epigenome and its effects on gene expression. We will then test these correlations by targeted modification of the epigenome using a modified CRISPR system. This system for targeted modification of the epigenome will also be explored as a system for epigenetic modification in crop plant improvement.