Current Evidence Does Not Support Substantial DNA Cytosine Methylation in Myxosporean Genomes

Abstract Some invertebrate lineages are known to have lost their ability to methylate DNA cytosines. These lineages have often lost the enzymes responsible for cytosine methylation (DNMT1 and 3), as well as other enzymes that interact with methylated cytosines. Previous studies showed that myxosporeans, a group of parasitic cnidarians, lack recognisable DNMTs and other canonical methylation enzymes, and observed a lack of cytosine methylation in two myxosporeans using whole genome bisulfite sequencing. Recently, a study of five myxosporean samples using Oxford Nanopore Technologies (ONT) sequencing reported high levels of methylation in two samples and low levels in the others and the presence of methylation-related genes in their assemblies. Here, we provide evidence that the two assemblies for which high levels of methylation were reported are heavily contaminated with fish-host sequences. In the other three samples, the lower apparent methylation signal resembles that observed in unmethylated negative controls and can be explained by ONT-background noise. Moreover, the CpG observed-to-expected ratios of the five assemblies differ from those of well-curated myxosporean assemblies. Additionally, the reported methylation profiles do not exhibit the typical distribution seen in invertebrates, possibly reflecting the presence of host contaminants. Finally, we found that most of the genes analysed are not canonical components of cytosine methylation pathways, and that the few candidate methylation-related genes reported in myxosporeans are more consistent with bacterial contamination than with genuine myxosporean genes. Together, these observations underscore important caveats in inferring the presence or absence of cytosine methylation in parasite genomes, particularly when using emerging technologies, and suggest that there is currently no robust evidence for substantial cytosine methylation in myxosporeans.