Characterization of the enzyme for 5-hydroxymethyluridine production and its role in silencing transposable elements in dinoflagellates

Dinoflagellate chromosomes are extraordinary, as their organization is independent of architectural nucleosomes unlike typical eukaryotes and shows a cholesteric liquid crystal state. 5-hydroxymethyluridine (5hmU) is present at unusually high levels and its function remains an enigma in dinoflagellates chromosomal DNA for several decades. Here, we demonstrate that 5hmU contents vary among different dinoflagellates and are generated through thymidine hydroxylation. Importantly, we identified the enzyme, which is a putative dinoflagellate TET/JBP homologue, catalyzing 5hmU production using both in vivo and in vitro biochemical assays. Based on the near-chromosomal level genome assembly of dinoflagellate Amphidinium carterae, we depicted a comprehensive 5hmU landscape and found that 5hmU loci are significantly enriched in repeat elements. Moreover, inhibition of 5hmU via dioxygenase inhibitor leads to transcriptional activation of 5hmU-marked transposable elements (TEs), implying that 5hmU appears to serve as an epigenetic mark for silencing transposon. Together, our results revealed the biogenesis, genome-wide landscape and molecular function of dinoflagellate 5hmU, providing mechanistic insight into the function of this enigmatic DNA mark. In our study, we successfully obtained the near chromosome-scale genome assembly of A. carterae based on a combination of genome survey approach, long-read sequencing through Pacbio sequel II platform with the circular consensus sequencing (CCS) mode, and Hi-C sequencing data. For A. carterae genome annotation, we also performed transcriptome sequencing using both short-read RNA-seq (DNBSEQ PE150) and long-read sequencing (PacBio Sequel platform). To map the genome-wide distribution of 5hmU sites in A. carterae, we utilized an anti-5hmU antibody-based DNA immunoprecipitation method. To examine the effects of altered 5hmU levels on repeat elements, we treated A. carterae cells with and without 5hmU synthase inhibitor, 2,4-pyridinedicarboxylic acid (2-4P), followed by directional RNA sequencing to reveal their transcriptional change. Additionally, we conducted Whole Genome Bisulfite Sequencing (WGBS) to investigate the correlation between 5hmU and genome-wide 5mC loci in A. carterae. All treatments in this study have at least two biological replicates.