Marine DNA methylation patterns are associated with microbial community composition and inform virus-host dynamics

Abstract

Background: DNA methylation in prokaryotes is involved in many different cellular processes including cell cycle regulation and defense against viruses. To date, most prokaryotic methylation systems have been studied in culturable microorganisms, resulting in a limited understanding of DNA methylation from a microbial ecology perspective. Here, we analyze the distribution patterns of several microbial epigenetics marks in the ocean microbiome through genome-centric metagenomics across all domains of life.

Results: We reconstructed 15,056 viral, 252 prokaryotic, 56 giant viral, and 6 eukaryotic metagenome-assembled genomes from northwest Pacific Ocean seawater samples using short- and long-read sequencing approaches. These metagenome-derived genomes mostly represented novel taxa, and recruited a majority of reads. Thanks to single-molecule real-time (SMRT) sequencing technology, base modification could also be detected for these genomes. This showed that DNA methylation can readily be detected across dominant oceanic bacterial, archaeal, and viral populations, and microbial epigenetic changes correlate with population differentiation. Furthermore, our genome-wide epigenetic analysis of Pelagibacter suggests that GANTC, a DNA methyltransferase target motif, is related to the cell cycle and is affected by environmental conditions. Yet, the presence of this motif also partitions the phylogeny of the Pelagibacter phages, possibly hinting at a competitive co-evolutionary history and multiple effects of a single methylation mark.

Conclusions: Overall, this study elucidates that DNA methylation patterns are associated with ecological changes and virus-host dynamics in the ocean microbiome. Video Abstract.

Keywords: DNA methylation; Host–phage; Long-read sequencing; Metagenome-assembled genome; Ocean microbiome.

Fig. 1

Meta-epigenome analysis scheme of ocean surface samples. A schematic overview of meta-epigenomics. Meta-epigenomics using genome-centric metagenomics from the binning approach of short- and long-read assemblies, followed by identifying the epigenetic signals of genomes from long-read mapping

Fig. 2

Phylogenetic tree of MAGs obtained from SHIPPO. a A phylogenetic tree of prokaryotic SHIPPO MAGs using core genes from Phylophlan2. A total of 252 MAGs were obtained; each bar outside the tree represents the number of methyltransferase (MTase) genes present in each MAG. b The distribution of restriction enzyme and MTase types from SHIPPO MAGs across kingdoms. c Prokaryotic SHIPPO MAGs were compared against genomes from Tara Oceans (TARA) and Global Ocean Reference Genomes Tropics (GORG-Tropics) datasets using FastANI. d The number of genes associated with the restriction-modification (RM) system is plotted against the genome size for each ocean microbiome MAG (SHIPPO MAG, TARA, and GORG-Tropics). Points are shaped depending on the type of the complete and orphan RM system. MAG: metagenome-assembled genome; SHIPPO: Shipborne Pole-to-Pole Observations

Fig. 3

Meta-epigenomic profile of MAGs across all sampling stations. a, Principal coordinate analysis (PCoA) clustering by the 5-mer methylation features of Shipborne Pole-to-Pole Observations (SHIPPO) MAGs based on Kulczynski dissimilarity. Each point represents each species-level MAG in each sample. The black-dashed circles represent family-level clusters of MAGs across samples. The colored-solid lines represent species-level clusters belonging to Pelagibacteraceae across samples. b The maximum methylation ratios of motifs are represented in each MAG at the family level; the highest methylation value among all sample sites is colorized. c Population differentiation versus methylome across sampling stations. For the most prevalent Shipborne Pole-to-Pole Observations (SHIPPO) MAGs, scatterplots show the relationship of 5-mer methylome dissimilarity based on Bray–Curtis and population differentiation by sampling distance. MAG: metagenome-assembled genome (eukaryotic: eukMAG, prokaryotic: proMAG, viral MAG: vMAG, giant viral MAG: gvMAG); F_ST: fixation index

Fig. 4

Genome-wide epigenetic analysis of Pelagibacter MAG across samples. a The UpSet plot compares GANTC methylation at each genome position on a Pelagibacter MAG (SHIPPO_PRO_33) across samples. More than 0.5 of the methylation fractions were considered methylated at each genome position; the color of each bar depends on the genic (G: green), intergenic (I: navy), and regulatory region (R: orange). The column bar indicates the intersection of the number of methylation positions on the MAG across samples. The left bar represents the total number of methylated positions on the MAG for each sample. b Principal coordinate analysis (PCoA) clustering by SNMV and SNV based on the Bray–Curtis distance on overlapped regions for all samples. c A model of the methylation pattern according to the cell-cycle progression in Alphaproteobacteria. d The methylation fraction comparisons of the GANTC motif between genomic regions of ori (replication origin), ter (replication terminus), and other regions. e The genome-wide distribution of methylated fractions for the GANTC motif indicates the trend of cell-cycle progress throughout the genome. MAG: metagenome-assembled genome; SHIPPO: Shipborne Pole-to-Pole Observations; SNMV: single nucleotide methylation variation; SNV: single nucleotide variant

Fig. 5

DNA methylation of the viral genome. a The methylome of 83 prevalent viral Shipborne Pole-to-Pole Observations (SHIPPO) of metagenome-assembled genomes (vMAGs) is indicated by the heatmap with their phylogeny from genome similarities. The star represents vMAGs in c. b Phylogenetic comparison of the 14 vMAGs harboring MTase and its MTase genes. c Changes in vMAG methylation profiles were measured at three sampling stations. Circles represent the methylation ratio of each motif, and bars represent the read mapping breadth of viral genomes with 10× depth

Fig. 6

Genomic characterization between Pelagibacter and pelagiphage according to the GANTC motif. a Distribution of GANTC motif methylation in both strands of Cand. Pelagibacter Giovannoni NP1 genome. The inner blue and red bars indicate the coding sequence region of the genome. The methylation fraction bar for each strand is red for values ≥ 0.9, orange for ≥ 0.6, and bright yellow for ≥ 0.3. The prophage region of the genome is highlighted by a purple bar from Pleška et al. [26]. The GANTC motif depletion region is marked through a multiscale signal representation (MSR) analysis. b, Phylogenetic tree of pelagiphages of the head–tail connector protein. Red boxes represent the GANTC depletion subgroups of pelagiphage. c The GANTC motif density comparisons between genomes of Pelagibacter and pelagiphage