Metabolomic and Transcriptomic rewiring between virus resistance and susceptibility in Ostreococcus mediterraneus

The genus Ostreococcus, comprising picoeukaryotic unicellular algae, plays a key role in many coastal marine ecosystems. These phytoplankton are infected by lytic double-stranded DNA prasinoviruses against whom they have the capacity to reliably evolve stable antiviral resistance. Regulation of virus resistance and susceptibility is hypothesized to arise through a phenotypic switch, as resistant cell lines can be isolated from susceptible lines after viral exposure and vice versa. To elucidate the molecular mechanisms underlying this resistance, we integrated untargeted metabolomic analyses and transcriptomics in virus-resistant and virus-susceptible lines of O. mediterraneus. Transcriptomic analyses corroborated previous findings in O. tauri, revealing that most gene expression changes were concentrated on a single chromosome. Specifically, a [~]530 kb region was over-transcribed in virus-resistant lines, while a distinct [~]110 kb region was over-expressed in virus-susceptible lines. Comparative metabolomics identified several oxidized galactolipids and oxidized sterols as biomarkers of the susceptible phenotype, while fewer biomarkers were identified for the resistant phenotype. By integrating transcriptomic and metabolomic signatures--focusing on the expression of genes within biosynthetic pathways linked to these metabolite biomarkers--we uncovered candidate molecular mechanisms underlying the cellular physiology of susceptible versus resistant phenotypes. Authors SummaryViruses are the most abundant biological entities in the ocean, shaping the dynamics of phytoplankton communities that underpin marine food webs. Ostreococcus, one of the smallest photosynthetic eukaryotes, is frequently infected by highly abundant prasinoviruses in its natural environment. In this study, we aimed to understand the physiological bases of virus-resistant and susceptible Ostreococcus mediterraneus lines. We compared gene expression and untargeted metabolite profiles between a resistant line and a susceptible line derived from it. The metabolic profiles were much more variable between replicates than the transcriptomic profiles and the most differentially expressed genes did not include those involved in the biosynthetic pathways of the metabolite biomarkers identified. We estimated the congruence between the metabolomes and transcriptomes as the percent of relative expression changes fitting to the relative change in the metabolite. This study provides evidence of the subtle links between gene expression and metabolomic signatures and the importance of integrating multiple levels of cellular processes.