Project description:DNA methylation predominantly occurs at CG dinucleotides in vertebrate genomes, however, non-CG methylation (mCH) is also detectable in vertebrate tissues, most notably in the nervous system. In mammalian brains, it is well established that: i) mCH is targeted to CAC trinucleotides by DNMT3A, ii) enriched in gene bodies and repetitive elements, and iii) associated with transcriptional repression. However, the possible conservation of these mCH features in zebrafish is largely unexplored and has yet to be functionally demonstrated. In this study, we analyse the transcriptomes (RNA-seq) and methylome (RRBS) of developing zebrafish larvae (1-6 weeks) and adult brain (6 month). We additionally elucidate a role for dnmt3aa/dnmt3ab in mCH deposition via CRISP/CAS9 KO and WGBS of 4 week old brains
Project description:Two-thirds of gene promoters in mammals are associated with regions of non-methylated DNA, called CpG islands (CGIs), which counteract the repressive effects of DNA methylation. In lower vertebrates, computational CGI predictions often reside away from gene promoters, suggesting a major divergence in gene promoter architecture across vertebrates. By experimentally identifying non-methylated DNA in the genomes of seven diverse vertebrates, we instead reveal that non-methylated islands (NMIs) of DNA are a central feature of vertebrate gene promoters. Furthermore, NMIs are present at orthologous genes across vast evolutionary distances, revealing a surprising level of conservation in this epigenetic feature. By profiling NMIs in different tissues and developmental stages we uncover a unifying set of features that are central to the function of NMIs in vertebrates. Together these findings demonstrate an ancient logic for NMI usage at gene promoters and reveal an unprecedented level of epigenetic conservation across vertebrate evolution. Bio-CAP was used to identify non-methylated regions of the genome in seven diverse vertebrates (human, mouse, platypus, chicken, lizard, frog and zebrafish) across a number of tissues.
Project description:Positive-strand RNA viruses of the order Nidovirales have the largest known RNA genomes of vertebrate and invertebrate viruses with 36.7 and 41.1 kb, respectively. The acquisition of a proofreading exoribonuclease (ExoN) locus by an ancestral nidovirus enabled crossing of the 20 kb barrier. Other factors constraining genome expansions in nidoviruses remain poorly defined. Here, we assemble 76 genome sequences of invertebrate nidoviruses from >500.000 published transcriptome experiments and triple the number of known nidoviruses with >36 kb genomes, including the largest known 64 kb RNA genome. Many of the novel viral lineages acquired putative enzymatic domains that were inserted in open reading frame (ORF) 1a and ORF1b or equivalent regions and may constitute cofactors of the viral replicase or modulate infection otherwise. We classify multi-cistronic ExoN-encoding nidoviruses into seven groups and four subgroups, according to canonical and non-canonical modes of viral polymerase expression by ribosomes and genomic organization (reModes). The largest group employing the canonical reMode comprises invertebrate and vertebrate nidoviruses, including coronaviruses, with genomes ranging from 20 to 36 kb. Six groups with non-canonical reModes include giant invertebrate nidoviruses with 31 to 64 kb genomes. Among them are viruses with segmented genomes and viruses utilizing dual ribosomal frameshifting that we validate experimentally. Moreover, polyprotein length and genome size in nidoviruses show reMode- and host phylum-dependent relationships. We demonstrate that the largest polyproteins in nidoviruses may be close to an upper limit that we hypothesize to be determined by the host-inherent translation fidelity, further constraining nidovirus genome size. Thus, expansion of giant RNA virus genomes, the vertebrate/invertebrate host division, the control of viral replicase expression, and translation fidelity are interconnected.
Project description:Vertebrates have highly methylated genomes at CpG positions while most invertebrates have sparsely methylated genomes. Therefore, hypermethylation is considered a major innovation that shaped the genome and the regulatory roles of DNA methylation in vertebrates. However, here we report that the marine sponge Amphimedon queenslandica, belonging to one of the earliest branching animal lineages, has evolved a hypermethylated genome with remarkable similarities to that of a vertebrate. Despite major differences in genome size and architecture, independent acquisition of hypermethylation reveal common distribution patterns and repercussions for genome regulation between both lineages. Genome wide depletion of CpGs is counterbalanced by CpG enrichment at unmethylated promoters, mirroring CpG islands. Furthermore, a subset of CpG-bearing transcription factor motifs are enriched at Amphimedon unmethylated promoters. We find that the animal-specific transcription factor NRF has conserved methyl-sensitivity over 700 million years, indicating an ancient cross-talk between transcription factors and DNA methylation. Finally, the sponge shows vertebrate-like levels of 5-hydroxymethylcytosine, the oxidative derivative of cytosine methylation involved in active demethylation. Hydroxymethylation is concentrated in regions that are enriched for transcription factor motifs and show developmentally dynamic demethylation. Together, these findings push back the links between DNA methylation and its regulatory roles to the early steps of animal evolution. Thus, the Amphimedon methylome challenges the prior hypotheses about the origins of vertebrate genome hypermethylation and its implications for regulatory complexity.