Metatranscriptomes-based sequence similarity networks uncover genetic signatures within parasitic freshwater microbial eukaryotes

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Microbial eukaryotes play a crucial role in biochemical cycles and aquatic trophic food webs. Their taxonomic and functional diversity are increasingly well described due to recent advances in sequencing technologies. However, the vast amount of data produced by -omics approaches require data-driven methodologies to make predictions about these microorganisms’ role within ecosystems. Using metatranscriptomics data, we employed a sequence similarity network-based approach to explore the metabolic specificities of microbial eukaryotes with different trophic modes in a freshwater ecosystem (Lake Pavin, France). A total of 2,165,106 proteins were clustered in connected components enabling analysis of a great number of sequences without any references in public databases. This approach coupled with the use of an in-house trophic modes database improved the number of proteins considered by 42%. Our study confirmed the versatility of mixotrophic metabolisms with a large number of shared protein families among mixotrophic and phototrophic microorganisms as well as mixotrophic and heterotrophic microorganisms. Genetic similarities in proteins of saprotrophs and parasites also suggest that fungi-like organisms from Lake Pavin, such as Chytridiomycota and Oomycetes, exhibit a wide range of lifestyles, influenced by their degree of dependence on a host. This plasticity may occur at a fine taxonomic level (e.g., species level) and likely within a single organism in response to environmental parameters. While we observed a relative functional redundancy of primary metabolisms (e.g., amino acid and carbohydrate metabolism) nearly 130,000 protein families appeared to be trophic mode-specific. We found a particular specificity in obligate parasite-related Specific Protein Clusters, underscoring a high degree of specialization in these organisms. Although no universal marker for parasitism was identified, candidate genes can be proposed at a fine taxonomic scale. We notably provide several protein families that could serve as keys to understanding host-parasite interactions representing pathogenicity factors (e.g., involved in hijacking host resources, or associated with immune evasion mechanisms). All these protein families could offer valuable insights for developing antiparasitic treatments in health and economic contexts.

This publication has 96 references indexed in Scilit: