Deciphering the genetic basis of phytoplankton traits through genome-wide association studies

Abstract Recently, an inventory of genes in phytoplankton was conducted through expeditions such as TARA Oceans. Approximately 1.5 million genes were identified, of which at least three-quarters have unknown function. Presently, a several research programmes are engaged in the sequencing of marine biodiversity, resulting in a rapid expansion of genomic databases. Access to the genomic sequences of these organisms will soon be readily accessible to the scientific community. Although analysing this data is promising, the characterization of genes or genomes, on the other hand, is progressing very slowly and remains a major challenge for scientists. The aim of this study was to use GWAS approaches to decipher genomic loci without a priori assumptions. The microalga Tisochrysis lutea was selected as a case study due to its economic importance and the extensive knowledge accumulated over the years. Particular attention was paid to pigment and lipid metabolism due to their high commercial value. To implement the GWAS approach, a collection of algal lineages was established (100 lineages) from available polyclonal strains (15 strains). This collection was then phenotyped under two different culture conditions. Of the 31 phenotypic traits investigated, 18 met the requirements for GWAS analysis. Concurrently, each algal lineage was genotyped by whole genome sequencing to inventory all genetic polymorphisms. A mixed model was applied, revealing 13 significant associations between phenotypic traits and alleles. These associations highlight previously unsuspected genomic loci that play a major role in pigment or lipid content. Genes identified at these loci may have a direct or indirect role in these metabolic pathways. Nevertheless, elucidating the molecular mechanisms of the associated genes remains limited without the implementation of functional approaches. Despite the complexity of the process, we conclude that the GWAS approach was effective for deciphering phytoplankton genomes, particularly for quantitative traits of interest. Ideally, this approach should be combined with other functional methods to progressively decode marine genomes.