The Role of Chromatin Modifications in the Evolution of Giant Plant Genomes

Abstract

Angiosperm genome sizes (GS) range ~2400-fold and comprise genes and their regulatory regions, repeats, semi-degraded repeats, and 'dark matter'. The latter represents repeats so degraded that they can no longer be recognised as repetitive. In exploring whether the histone modifications associated with chromatin packaging of these contrasting genomic components are conserved across the diversity of GS in angiosperms, we compared immunocytochemistry data for two species whose GS differ ~286-fold. We compared published data for Arabidopsis thaliana with a small genome (GS = 157 Mbp/1C) with newly generated data from Fritillaria imperialis, which has a giant genome (GS = 45,000 Mbp/1C). We compared the distributions of the following histone marks: H3K4me1, H3K4me2, H3K9me1, H3K9me2, H3K9me3, H3K27me1, H3K27me2, and H3K27me3. Assuming these histone marks are associated with the same genomic features across all species, irrespective of GS, our comparative analysis enables us to suggest that while H3K4me1 and H3K4me2 methylation identifies genic DNA, H3K9me3 and H3K27me3 marks are associated with 'dark matter', H3K9me1 and H3K27me1 mark highly homogeneous repeats, and H3K9me2 and H3K27me2 mark semi-degraded repeats. The results have implications for our understanding of epigenetic profiles, chromatin packaging and the divergence of genomes, and highlight contrasting organizations of the chromatin within the nucleus depending on GS itself.

Keywords: chromatin; dark matter; epigenetics; giant genomes; histone modifications; immunocytochemistry.

Figure 1

Profile of the repeat landscape in Fritillaria imperialis generated via RepeatExplorer2 [20]. Larger coloured boxes in the legend are shown for Athila (~7.4%) and Tekay (~9.5%) Ty3/Gypsy elements and the EnSpm CACTA-like DNA transposon (~2.3%). Unknown repeat types are shown in grey and amount to 5.5%. Repeat abundances less than 0.05% are not shown.

Figure 2

Similarities and differences in the distributions of specific methylated histone H3 marks of (A) Fritillaria imperialis (this study) and (B) Arabidopsis thaliana (images of labelled nuclei taken from Fuchs and Schubert, [21]), which differ ~286-fold in genome size, together with (C) the interpretation of the histone methylation targets. Immunolabelling with antibodies against specific methylation marks (me1, me2, and me3) on lysines K4, K9, and K27 of histone H3 are visible as pink fluorescent signal on interphase nuclei that were counterstained with DAPI (blue fluorescent signal). Scale bars are shown, but to better illustrate the differences in nuclear size between A. thaliana and F. imperialis, a nucleus of each species is shown at the same magnification in (A) labelled for H3K9me1.

Figure 3

Diagrammatic summary of the distribution of different genomic domains in Fritillaria imperialis and Arabidopsis thaliana differing by ~286-fold in genome size. Genes in F. imperialis appear to be widely dispersed but comprise only a tiny fraction of the genome (i.e., ~0.07%), and it is unknown if they are localised to any particular domain. Note that the purple ‘spots’ in F. imperialis predicted to contain semi-degraded repeats and ‘dark matter’ are similar in size to the entire A. thaliana nucleus (see also Figure 2A).