Engineering natural microbiomes toward enhanced bioremediation by microbiome modeling

Microbes are essential for various biogeochemical cycles, including nutrient metabolism, agriculture, food fermentation, electromechanics, biofuels, and pollutant degradation. Synthetic microbiomes, based on interacting relationships within microbiomes, offer improved efficiency compared to single strains or natural microbiomes, promising applications in various fields. These synthetic microbiomes offer a strategy to establish complex metabolic functions by combining the metabolic capacities of multiple strains, overcoming limitations of single strains and enabling the sharing of metabolic burdens. Despite efforts in constructing synthetic strains, research on synthetic microbiomes is still in its infancy, lacking practically applicable principles and tools for natural microbiome engineering. Bottom-up and top-down strategies have been proposed, but practical application remains a challenge. This study aims to address this gap by developing a combined top-down and bottom-up framework for engineering natural microbiomes for enhanced bioremediation of herbicide-contaminated soils.

The existing literature highlights the importance of microbes in various biogeochemical cycles and the potential of synthetic microbiomes. Studies have demonstrated the successful construction of synthetic strains for specific functions, but the engineering of complex microbiomes remains challenging. The literature emphasizes the lack of practical principles and tools for effectively engineering natural microbiomes. Bottom-up and top-down strategies have been explored individually, but a combined approach is lacking. Previous research shows the potential of microbial consortia for bioremediation, but the identification of key species and the understanding of metabolic interactions are often limited. This research builds on existing knowledge by integrating both top-down and bottom-up approaches to create a more effective and comprehensive framework.

This study used a combined top-down and bottom-up approach. The top-down phase involved engineering natural microbiomes from three different soil types (red, yellow, and purple) using herbicide application (BO and DBH) and inoculating with different combinations of herbicide-degrading strains (single strains, synergistic consortia, and competitive consortia). The impact of herbicide application and inoculation was assessed by analyzing the dynamic development of the functional microbiome and its bioremediation ability. Repeated high-dose inoculations were found to be most effective. The bottom-up phase involved identifying keystone species from the engineered microbiomes using a combination of 16S rRNA gene amplicon sequencing, metagenomics, and strain isolation. LEfSe analysis was used to identify genera with significantly different abundances in treated versus untreated microbiomes. A total of 18 keystone species were identified. These keystone species were then used to construct simplified microbiomes. A newly developed microbiome modeling framework, SuperCC, was used to simulate the performance of different combinations of keystone species under various conditions. SuperCC is a scalable modeling framework that integrates single-species models into a multi-compartment model, simulating metabolic flux distributions in microbiomes. The performance of the simplified microbiomes was experimentally validated using pot experiments and various analyses, including HPLC for measuring pollutant degradation, LC-MS for detecting exchanged metabolites, 13C DNA-SIP, and RNA-seq for transcriptomic profiling of key species. Computational cell design was also explored by identifying essential metabolic reactions from the functional microbiomes and transferring them to target cells. Statistical analyses such as ANOVA, random forest, PCA, and linear regression were used throughout the study to analyze the data.

The study demonstrates that herbicide application and herbicide-degrader inoculation effectively enhance bioremediation capabilities in different soil microbiomes. Microbiome reassembly occurs at both taxonomic and functional levels, converging toward a functional microbiome with improved pollutant-degrading efficiency regardless of initial microbiome composition. The study successfully identifies 18 keystone species crucial for enhanced bioremediation. The SuperCC modeling framework accurately predicts the performance of simplified microbiomes with different combinations of keystone species, allowing for optimization of the microbiome composition. The metabolic interactions among keystone species, particularly mutualistic relationships, are identified and experimentally validated through various techniques such as LC-MS, 13C DNA-SIP, and RNA-seq. The study demonstrates that the exchanged metabolites between keystone strains significantly enhance growth and degradation. Computational design of synthetic cells, by integrating essential metabolic reactions from the functional microbiomes, shows promise in achieving enhanced bioremediation. The convergent succession of different microbiomes, regardless of initial conditions, suggests a universal principle underlying functional microbiome assembly under selective pressure.

The findings demonstrate the feasibility of engineering natural microbiomes for improved bioremediation through a combined top-down and bottom-up approach. The successful application of SuperCC highlights the power of microbiome modeling in optimizing microbiome composition and predicting performance. The identification of mutualistic interactions among keystone species emphasizes the importance of considering these complex interactions in designing effective synthetic microbiomes. The computational design of synthetic cells, mimicking the metabolic networks of functional microbiomes, offers a novel approach for creating efficient and stable bioremediation systems. This approach offers an alternative to traditional trial-and-error methods, potentially accelerating the development of effective bioremediation strategies. Future research should focus on expanding the range of pollutants and soil types, investigating the long-term stability of engineered microbiomes, and exploring applications in different environmental settings.

This study presents a novel framework for engineering natural microbiomes for enhanced bioremediation. The combined top-down and bottom-up approach, coupled with the SuperCC modeling framework, enables efficient identification of keystone species and optimization of microbiome composition. The findings emphasize the importance of metabolic interactions and provide practical guidance for designing effective synthetic microbiomes for various biotechnological applications. Future research should focus on validating these findings across a wider range of environmental conditions and pollutants.

The study focused on a limited set of herbicides and soil types. While the results suggest general principles, further research is needed to validate these findings across a broader range of conditions. The SuperCC model, while powerful, relies on assumptions about metabolic interactions and may not capture all the complexities of real-world microbiomes. The experimental validation, while extensive, may not completely account for all possible interactions within the complex soil environment. Long-term stability and ecological impacts of engineered microbiomes require further investigation.