Extensive epigenetic reprogramming during the life cycle of Marchantia polymorpha

Abstract

Background: In plants, the existence and possible role of epigenetic reprogramming has been questioned because of the occurrence of stably inherited epialleles. Evidence suggests that epigenetic reprogramming does occur during land plant reproduction, but there is little consensus on the generality and extent of epigenetic reprogramming in plants. We studied DNA methylation dynamics during the life cycle of the liverwort Marchantia polymorpha. We isolated thalli and meristems from male and female gametophytes, archegonia, antherozoids, as well as sporophytes at early and late developmental stages, and compared their DNA methylation profiles.

Results: Of all cytosines tested for differential DNA methylation, 42% vary significantly in their methylation pattern throughout the life cycle. However, the differences are limited to few comparisons between specific stages of the life cycle and suggest four major epigenetic states specific to sporophytes, vegetative gametophytes, antherozoids, and archegonia. Further analyses indicated clear differences in the mechanisms underlying reprogramming in the gametophytic and sporophytic generations, which are paralleled by differences in the expression of genes involved in DNA methylation. Differentially methylated cytosines with a gain in methylation in antherozoids and archegonia are enriched in the CG and CHG contexts, as well as in gene bodies and gene flanking regions. In contrast, gain of DNA methylation during sporophyte development is mostly limited to the CHH context, LTR retrotransposons, DNA transposons, and repeats.

Conclusion: We conclude that epigenetic reprogramming occurs at least twice during the life cycle of M. polymorpha and that the underlying mechanisms are likely different between the two events.

Keywords: Bisulfite sequencing; DNA methylation; Epigenetics; Life cycle; Liverwort; Marchantia polymorpha; Reprogramming; Tissue specificity.

Fig. 1

Illustration of the life cycle of M. polymorpha. Land plants have a more complex life cycle than animals, alternating between two multicellular, heteromorphic generations: the sporophyte and the gametophyte. In bryophytes (mosses, hornworts, and liverworts), the gametophytes constitute the dominant generation while the sporophyte is a small and simple structure, whose development depends on the female gametophyte. The male gametophyte forms antheridiophores, the reproductive structures that harbor the antheridia with the antherozoids (sperm). The female gametophyte forms archegoniophores, the reproductive structures that harbor the archegonia, each of which contains a single egg cell. During sexual reproduction, the antherozoids are released from the antheridia and swim towards the archegonia on the female gametophyte. The sporophyte is formed upon fertilization and remains attached to the archegoniophore during its entire development. Spores are formed in the sporophyte through meiosis and are finally released. The spores germinate and develop into either male (with Y chromosome) or female (with X chromosome) gametophytes, thereby concluding the life cycle. Both, male and female gametophytes are capable of asexual reproduction through the formation of gemmae in the gemma cups [57]. Numbers in magenta mark the tissues isolated for this study: (1/2) thallus of the female/male gametophyte, respectively, (3/4) apical notch of the female/male gametophyte, respectively, (5) archegonia, (6) antherozoids, (7) early sporophyte, and (8) late sporophyte

Fig. 2

a DNA methylation levels in percent at individual cytosines located in autosomes across all or within each individual sequence context (CG, CHG, CHH) for each tissue type used in this study shown as violin plots (see Additional file 1: Figure S1 for data from individual replicates). The horizontal black bars correspond to the means (see Additional file 1: Tables S2 and S3). b Average DNA methylation levels in percent for each sequence context, genomic feature, and tissue type shown as a heatmap (see Additional file 1: Table S4, S5, S6, and S7). US/DS upstream/downstream of a gene, UTR untranslated region, snRNA small nucleolar RNA, rRNA ribosomal RNA, sat satellite repeat, telo. sat telomeric satellite repeats, simple simple repeats, ukn. unknown/unclassified repeats, SINE short interspersed nuclear elements, LINE long interspersed nuclear elements, RC rolling circle transposon, LTR-TE retrotransposon with long terminal repeats, DNA-TE DNA transposon, uncl. not classified into a subfamily. *Note that for the “female gametes,” we sampled archegonia, which are entire gametangia that harbor the egg cells

Fig. 3

a DNA methylation levels in percent at individual cytosines located in sex chromosomes across all or within each individual sequence context (CG, CHG, CHH) for each tissue type used in this study shown as violin plots. The horizontal black bars correspond to the means (see Additional file 1: Table S9). Female/male gametophytes and gametes only contain the X/Y chromosome. Sporophytes contain both, X and Y, sex chromosomes. b Difference in the average DNA methylation level between the individual sex chromosomes and the autosomes for each sequence and genomic feature context. Fields marked with an asterisk depict comparisons that were statistically significant (two-sided t-test, adjusted for multiple testing, FDR < 0.05). Gray fields depict cases in which the given genomic feature is not present on the sex chromosome. US/DS upstream/downstream of a gene, UTR untranslated region, snRNA small nucleolar RNA, rRNA ribosomal RNA, sat satellite repeat, telo. sat telomeric satellite repeats, simple simple repeats, ukn. unknown/unclassified repeats, SINE short interspersed nuclear elements, LINE long interspersed nuclear elements, RC rolling circle transposon, LTR-TE retrotransposon with long terminal repeats, DNA-TE DNA transposon, uncl. not classified into a subfamily. *Note that for the “female gametes,” we sampled archegonia, which are entire gametangia that harbor the egg cells

Fig. 4

Schematic representation of the comparisons performed during the analysis of differential cytosine methylation. Variation in DNA methylation at each individual cytosine was analyzed with a linear model according to a design with a single factor comprising all different experimental groups [14]. Specific groups were then compared with linear contrasts. Percentages of cytosines with significant (FDR < 0.001) differences in DNA methylation are given for each comparison (total number of cytosines tested: 994,696). In addition, there is a histogram with the differences in DNA methylation at the individual DMCs for each comparison (the sizes are not proportional to the number of DMCs found in the comparison). The x-axis of the histogram ranges from – 100% to + 100% and the vertical line is at 0. The orientation of the x-axis is such that a higher density in the left/right side of the histogram corresponds to higher methylation in the group at the left/right side of the arrow depicting the comparison. For example, almost all DMCs identified in the comparison “gametophyte vs sporophyte” have increased methylation levels in the sporophytes compared to the gametophytes. See Additional file: Figure S4 for differences in DNA methylation for each individual sequence and genomic feature context

Fig. 5

Characterization of DMCs indicative of epigenetic reprogramming, i.e. DMCs that gained methylation in (a) antherozoids or archegonia compared to vegetative gametophytic tissues (309,834 DMCs), (b) early sporophytes compared to gametangia/gametes (24,073 DMCs), and (c) during sporophyte development (15,549 DMCs). Their distribution across the three sequence contexts (CG, CHG, and CHH) is shown as pie charts on the left. For each sequence context, the number of DMCs associated with a given genomic feature context is shown in a spider graph on the right. TE transposable element, LTR long terminal repeat, LINE long interspersed element, SINE short interspersed element, RC rolling circle transposon, US/DS upstream/downstream gene flanking regions, UTR untranslated region. “Repeats” included satellite, telomeric satellite, rRNA, snRNA, and simple repeats. “Repeats (ukn)” included unknown and unclassified repeats