(591c) Dynamic Genome-Scale Metabolic Modeling Suggests the Establishment of Mutualism without Co-Evolution within a Synthetic Microbiome

In the US, the biogas produced from anaerobic digestion (AD) of industrial, municipal and agricultural waste streams has immense potential as a renewable feedstock, with more than 7.8 M tone of CH₄ and over 14.4 M tone of CO₂ per year. However, the utilization of biogas is limited to heat and electricity production due to the presence of contaminants such as H₂S, NH₃, and volatile organic compounds (VOCs). The low value of biogas derived products (heat and electricity) has severely limited the installation of anaerobic digesters due to unfavorable return of investment. Therefore, to tap into the immense biogas potential from waste streams to produce valuable fuels and bioproducts, as well as to significantly reduce GHG emissions, effective biotechnologies that can operate at ambient pressure and temperature without requiring biogas cleaning/upgrading are urgently needed. In addition, the technologies that could co-utilize both CH₄ and CO₂ are especially preferred, as the simultaneous conversion can significantly reduce the equipment and operation cost.

To address this urgent need, we have proposed to use a synthetic microbiome, i.e., methanotroph-photoautotroph (M-P) coculture, as biocatalyst to achieve efficient, simultaneous conversion of both CH₄ and CO₂ for bioplastic production. In our prior research, we have demonstrated that the M-P coculture offers a highly effective platform for biogas valorization to fuels and chemicals, largely attributed to the metabolic coupling of methane oxidation and photosynthesis oxygenation. Microbial communities, including the synthetic M-P coculture, offer a number of advantages and hold great potential for future biotechnology development^1,2. However, the utilization of microbial communities for biotechnological applications have been limited. This is mainly due to: (1) lack of the tools to efficiently characterize the mixed culture; (2) highly complex dynamics of the mixed culture; (3) largely unknown interdependencies among different species.

To fully characterize the synthetic M-P coculture in real-time, we have developed a fast and easy experimental-computational (E-C) protocol based on the gas consumption rates, overall mass balance, and each organism’s growth stoichiometry. Through the E-C protocol, we can accurately determine the individual biomass concentrations, the uptake rate and production rate of O₂ and CO₂ for each species in real-time^3–5. The accuracy of the E-C protocol was experimentally validated to outperform flow cytometry based approach³. These real time measurements enable the development and validation of a semi-structured kinetic model to quantify the coculture effect on the cell growth rates⁶. Additionally, our experiments have confirmed that within a model M-P coculture, Methylomicrobium buryatense – Arthrospira platensis, there exist other emergent metabolic interactions (in addition to the exchange of the in situ produced O₂/CO₂) that significantly enhanced the growth of both species. To quantify the effect of these unknown emergent interactions on the growth of each partner, we developed a semi-structured kinetic model that can accurately predict the coculture growth under a wide range of conditions. Remarkably, the maximum cell growth rate for the methanotroph and cyanobacteria in the coculture showed 42% and 48% increase compared to those in the monocultures, respectively^7,8.

Many factors inherently complicate most mutualistic interactions; therefore, it can be difficult to isolate their components for specific studies and determine if they are truly contributing to the observed mutualism⁹. Genome-scale metabolic models (GEMs) offer a convenient and powerful tool for integrative analysis of different sources of data, including -omics data, as well as generating and testing various hypotheses regarding “metabolic links” within microbial consortia. To postulate potential molecular mechanisms responsible for the enhanced growth observed in the coculture, here we present the steady-state GEM we developed for the model M-P coculture using both SteadyCom¹⁰ and Microbiome Modeling Toolbox (MMT)¹¹.

One prerequisite for the development of the coculture GEM is refined GEMs for each individual species. Building upon our prior results for efficient GEM analysis and refinement, the system-identification based framework^12–14, we have validated and refined the available GEMs for M. buryatense and A. platensis. While several GEM modeling approaches have been developed for microbial communities, little emphasis has been placed on the need for actual and reliable constraints of these complex systems, especially when the behavior of these communities is dynamic. To address these challenges, we employ the validated semi-structured kinetic model to provide the cross-membrane fluxes, i.e., substrate pickup rates, product excretion rates and biomass growth rates, as constrains to the coculture GEM. These constraints not only allow the reduction of the feasible solution space of the GEM model, but also enables the simulation of the dynamic GEM. Because the gene-regulation is much faster than the bioreactor dynamics, we can assume that the cellular metabolism is always in a pseudo-steady-state. Therefore, we can use the substrate uptake rates predicted by the kinetic model as constraints to the GEM, and use the steady-state GEM to determine the metabolic flux profile under then given pseudo-steady state. In this way, by integrating the kinetic model with the steady-state GEM for the coculture, we were able to obtain the dynamic metabolic flux changes over time.

The coculture GEM was validated by comparing the model predicted individual cell growth rates with experimental measurements. Our extensive simulations under various in silico setups with the GEM consistently predicted the same top exchanged metabolites, as shown in Figure 1(a). It is also interesting to point out that the dynamic coculture GEM predicts the establishment of the mutualistic relationship between the methanotroph and cyanobacteria. Specifically, the dynamic GEM predicts an emergent N-exchange being established when M. buryatense and A. platensis are cocultured together. After the establishment of the mutualistic interaction, only the cyanobacteria uptakes nitrate, while the methanotroph consumes ammonia excreted by the cyanobacteria. The establishment of this mutualism can be observed clearly by comparing the methanotroph nitrate uptake in figure 1 (a) (at time = 8hrs, where nitrate uptake by the methanotroph is significant) and figure 1 (b) (at time =50 hours, where nitrate uptake by the methanotroph is zero). We are currently working on validating these predicted metabolic exchanges experimentally.

References

1- Bernstein HC, Carlson RP. Microbial consortia engineering for cellular factories: in vitro to in silico systems. Comput Struct Biotechnol J. 2012;3(4):e201210017.

2- Brenner K, You L, Arnold FH. Engineering microbial consortia: a new frontier in synthetic biology. Trends Biotechnol. 2008;26(9):483-489. http://www.sciencedirect.com/science/article/pii/S0167779908001716

3- Badr K, Whelan W, He QP, Wang J. Fast and easy quantitative characterization of methanotroph–photoautotroph cocultures. Biotechnol Bioeng. 2021;118(2):703-714.

4- Stone KA, Shah D, Kim MH, Roberts NRM, He QP, Wang J. A Novel Soft Sensor Approach for Estimating Individual Biomass in Mixed Cultures. Biotechnol Prog. 2017;33(2):347-354. doi:10.1002/btpr.2453

5- Stone KA, He QP, Wang J. Two Experimental Protocols for Accurate Measurement of Gas Component Uptake and Production Rates in Bioconversion Processes. Sci Rep. 2019;9(1):5899. doi:10.1038/s41598-019-42469-3

6- Badr K, Hilliard M, Roberts N, He QP, Wang J. Photoautotroph-Methanotroph Coculture–A Flexible Platform for Efficient Biological CO2-CH4 Co-utilization. IFAC-PapersOnLine. 2019;52(1):916-921.

7- Badr K, He QP, Wang J. Semi-structured kinetic modeling of a methanotroph-cyanobacteria coculture for biogas conversion. Bioresour Technol. Published online 2021:(submitted).

8- Roberts N, Hilliard M, Badr K, Whelan W, He QP, Wang J. A Novel Biological Platform for Simultaneous Biogas Conversion and Nutrient Recovery for Wastewater Treatment. In: 21st IFAC World Congress.

9- Dunham MJ. Synthetic ecology: a model system for cooperation. PNAS. 2007;104:1741–1742.

10- Chan SHJ, Simons MN, Maranas CD. SteadyCom: Predicting microbial abundances while ensuring community stability. PLoS Comput Biol. 2017;13(5):e1005539.

11- Baldini F, Heinken A, Heirendt L, Magnusdottir S, Fleming RMT, Thiele I. The Microbiome Modeling Toolbox: from microbial interactions to personalized microbial communities. Bioinformatics. 2019;35(13):2332-2334.

12- Damiani AL, He QP, Jeffries TW, Wang J. Comprehensive evaluation of two genome-scale metabolic network models for Scheffersomyces stipitis. Biotechnol Bioeng. 2015;112(6):1250-1262.

13- Hilliard M, Damiani A, He QP, Jeffries T, Wang J. Elucidating redox balance shift in Scheffersomyces stipitis’ fermentative metabolism using a modified genome-scale metabolic model. Microb Cell Fact. 2018;17(1):140.

14- Wang J, He QP, Damiani A, He QP, Wang J. A System Identification Based Framework for Genome-Scale Metabolic Model Validation and Refinement. In: Proceedings of Foundations of Systems Biology in Engineering 2015. ; 2017:13013-13018.