Abstract :
[en] Biological wastewater treatment is based on the use of microorganisms capable of intense metabolic activity that results in the removal of a large proportion of organic and inorganic contaminants. Given copious amounts of energy-dense organic molecules such as lipids accumulated by the microbial biomass, chemical energy may be directly harnessed from this for biofuel production. Here, lipid accumulating organism (LAO)-enriched microbial communities were studied using a molecular eco-systems biology approach. This involved the development of necessary methodologies including a new comprehensive biomolecular extraction method, yielding high-quality DNA, RNA, proteins and metabolites, as well as bioinformatic approaches for integrating and analysing the derived high-throughput genomic, transcriptomic, proteomic and metabolomic data. At the inception of the project, a full-scale wastewater treatment plant (WWTP) system with a strong presence of LAOs especially during winter months, i.e. the Schifflange WWTP (Esch-sur-Alzette, Luxembourg), was identified and selected for further study. 16S rRNA amplicon sequencing highlighted the presence of ubiquitous lipid accumulating bacteria closely related to Candidatus Microthrix parvicella which increase in abundance from autumn to winter over other highly abundant community members belonging to Alkanindiges spp. and Acinetobacter spp. In order to elucidate compositional, genetic and functional differences between autumn and winter LAO communities, a comparative integrative omic analysis was carried out on rationally selected autumn and winter LAO community samples. The results from metabolomic/lipidomic analyses between intra- and extracellular compartments support previous models of uptake of unprocessed long chain fatty acids (LCFAs) from the wastewater environment and their storage as triacyglycerols within LAOs. Furthermore, a tailored computational framework for the integration of multi-omic datasets into reconstructed community-wide metabolic networks and models was developed. The resulting networks provide overviews of functional capacity of the sampled LAO communities by incorporating gene copy numbers, transcript levels and protein frequency across the two studied environmental conditions. The identification of genes overexpressed, strongly associated with a specific season and/or possessing a high clustering coefficient suggests the existence of keystone genes, analogous to keystone species in species interaction networks. Examples of such keystone genes in the context of the LAO communities include genes coding for proteins involved in nitrogen and glycerolipid metabolism. The existence of such keystone genes opens up exciting possibilities for prediction and control strategies of microbial communities at the dawn of the field of Eco-Systems Biology.
