טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentShany Ofaim
SubjectThe Development of Computational Approaches for the
Modeling of Metabolic Functional-Division in
Bacterial Communities
DepartmentDepartment of Biotechnology and Food Engineering
Supervisors Full Professor Kashi Yechezkel
?
Full Thesis textFull thesis text - English Version


Abstract

Background: Microbial communities are abundant in nature and can metabolize myriad of compounds. The composition of the community together with its internal interactions can affect both the rate and the accumulation of interim and end products. Community dynamics can be viewed as dictated by a triangle of environment (secreted/other resources), community (metabolic-conversions repertoire), and function (excretion of altered forms). As advance technologies increased the need for computational analysis tools, metagenomics functional analysis has become of increasing interest. Here, I demonstrate the use of two network based approaches for the functional analysis of genomic & metagenomics data. These span the analysis of a single genome through a genome-scale metabolic model (quantitative) to the analysis of the functional potential of complex communities using metabolic networks approaches integrated with the novel application of ecological concepts (qualitative). The application of the qualitative approach was used to study the intricate associations of the complex rhizosphere and soil communities. Niche-specific metabolic-networks for metagenomics-based gene-catalogues derived from root and respective soil samples were constructed and assigned network-edges (metabolic-functions) with taxonomic annotations. Simulations of activity following removal or addition of enzymes specifically associated with distinct taxonomic groups allowed the prediction of species-specific positioning within the community food-web as well as predict potential metabolic exchanges leading to the co-production of complementary metabolites. The quantitative approach was presented as an Arthrobacer aurescens TC1 genome-scale metabolic model reconstruction as a predictive simulator for atrazine degradation optimization. 

Results: Using the metabolic networks approach, simulations of the removal of dominant key taxonomic groups resulted in group-specific and niche-dependent effects and differs between root and soil. Root-specific effects linked the utilization of specific plant exudates (e.g., flavonoids, organic acids) with specific taxonomic groups, pointing at the impact of each such compound as a determinant of the microbial community structure. Profiles of complementary metabolites suggest possible metabolic role behind observed co-occurrence patterns of bacterial combinations, for example through the production of nutrients that are in high demand in a specific niche. When considering the quantitative   approach, following the construction and curation of a genome scale metabolic model for Arthrobacer aurescens TC1, Constraint Based Simulations lead to prediction of growth yields on several carbon and nitrogen sources, focusing on atrazine and its interim degradation products. Further simulations, designed to test complementation between atrazine and a wide spectrum of metabolites have shown a positive effect on both atrazine degradation and biomass yield for the addition of glucose or riboflavin. Predictions for the effect of several media on growth and degradation rate were tested and validated in vitro.

Conclusions: Considering the volume of data produced in metagenomics studies, the developed approaches contribute to the current efforts for developing new conceptual approaches for the functional analyses of metagenomics. These allow the analysis of a wide spectrum of data complexity, from a single genome to a complex community. These integrated approaches produce testable predictions for process optimization or the initial investigation of species contribution to community function, function significance and structural shifts in complex bacterial communities.