Background
Negative impact of landfill gas to the environment
Landfill gas consists of an approximately 1:1 ratio of methane (CH4) to carbon dioxide (CO2). Methane in particular is responsible for a greater degree of global warming compared to carbon dioxide, and landfill sites contribute to 23% of nationwide annual greenhouse gas emissions.
As of 2020, landfills contribute to 23% of national methane emissions, of which 52% of the gas was used to generate low-carbon electricity, 17% to produce renewable natural gas and 30% was used at nearby facilities [1] . Through the Government of Canada’s Strengthened Climate Plan - A Healthy Environment and a Healthy Economy, the nation has planned to implement federal regulations to increase the number of landfills that collect their methane, while exploring biosolids and waste management infrastructure [2] .
How are landfill gasses processed today?
Currently, the Green Lane Landfill collects landfill gas, predominantly methane (CH4) and carbon dioxide (CO2), and directs it to a flaring system, converting it into CO2 before its release into the atmosphere [3] , emitting millions of cubic meters of CO2 [4] .
First Nation impacted by the landfill sites
A large number of Canada First Nation sites and Indigenous communities are located in the vicinity of landfill sites, which produce biogas through anaerobic decomposition.
Our Product
A biological machine for methane capturing at landfill sites
Our project aims to engineer bacteria capable of consuming methane and methane derivatives, such as methanol, formate, and formaldehyde, and implement this system in landfill sites.
1. Separation of CH4 and CO2 from Landfill Gas with Hardware
Why
Current solutions burn the methane acquired from landfill gas either directly to the air or as a form of energy production, both of which are carbon-dioxide emitting. The synthetic biology solution we pursue involves the respiration of methanol formed from landfill gas methane. To do this, the methane from landfill gas must be isolated and then oxidized to methanol. This methanol will serve as a carbon feedstock into methylotrophic strain bacteria. The bacteria that feed on the methanol must also be cultivated on an industrial scale to effectively remove methane from the landfill gas ecosystem.
Goal
Our hardware team’s project aims to fill this gap by designing a separation process for the landfill gas, a process for oxidizing the resultant methane gas, and a fermenter to cultivate the methylotrophic bacteria at scale.
Hardware Part 1: Separation
The composition of landfill gas is approximately 50% methane, 45% carbon dioxide, and 5% nitrogen. The separation is to take place across two stages; the first one to remove carbon dioxide from the landfill gas mixture, and the second to remove nitrogen from methane [5] . For each stage, we plan on sizing the membranes by building a model with MATLAB, and incorporating the results of the modeling into Aspen Plus. We also hope to prototype the carbon dioxide removal separation by building a downscaled model and demonstrating separation with nitrogen/carbon dioxide mixtures.
Hardware Part 2: Oxidation of methane to methanol
The difficulty of this process is that methane is usually only completely oxidized to carbon dioxide. Oxidizing to methanol is a more intricate process that needs precise reaction control. We plan on accomplishing this with a photochemical/electrochemical process with a catalyst. This process will be modeled and sized with Aspen Plus.
Hardware Part 3: Fermenter Construction
To design the fermenter, we needed to understand the context the fermenter will be placed in: where, to what size, etc. We plan on organizing a site visit to a landfill where we can learn to what scale we need to construct a plant for the fermenter. We will design and size the fermenter with Aspen Plus. We also hope to develop process controls for the fermenter to optimize bacteria growth parameters such as feed flow, temperature, pH, and stirring.
2. Constructing Improved Methylotrophic Strains
Developing a collection of methylotrophic microbes is essential for fundamental research in this field. Through a variety of experimental techniques, we aim to enhance the assimilation of single-carbon compounds in our engineered strains:
| Experimental Technique | Function |
|---|---|
| Overexpression | Introducing a recombinant plasmid into the chassis through cloning |
| CRISPR-Cas System Knockout | Knocking out undesired genes through gene editing and pathway engineering to optimize methylotrophic survival. We intend to use different CRISPR Cas systems to knockout different genes of interest. |
| Adaptive Evolution | Optimizing the fitness of the engineered strain by maintaining a stressor in the growth environment (methanol). |
Our Strain of Interest: T-B18 E. coli
T-B18 E. coil is E. coli BW25113 with frmA knocked out. It harbors pETM6_Ptrc_BsMdh_BmHps_BmPhi plasmid (Addgene plasmid #129104), and was adaptively evolved to grow efficiently on threonine and perform significant biosynthesis while employing methanol-derived carbon [6] T-B18 was a gift from Maciek R. Antoniewicz & Eleftherios Terry Papoutsakis (Addgene plasmid # 159446)
The RuMP pathway is an energetically favorable pathway that is compatible with the industrial E. coli strains [7, 8]; T-B18 is an evolved E. coli strain that assimilates methanol into its central carbon metabolism through the RuMP pathway. Pathway modifications that enhance RuMP cycle output are of key interest for maximizing biomass production.
Our Strain of Interest: C1-G-S-EVO P. Putida
The Reductive Glycine pathway (rGlyP) is a promising pathway for one carbon compound assimilation, especially for it allows formatotrophic growth. It requires only one ATP input and three reducing agents per linear pathway. We contacted Prof. Claassens for P. Putida strain utilizing rGlyP, who generously offered us strain C1-G-S-EVO described in the paper [9] .
Metabolic Modelling and Simulation Software
Why are computational models important for metabolic engineering?
Metabolism is a highly complex network of thousands of reactions, and it’s impossible to take into account the impact of each gene, metabolite, enzyme, or even culture media without a mathematical model. In order to investigate our strains of interest (T-B18 and C1-G-S-EVO P. Putida), we constructed genome-scale-metabolic models (GSMM) for those chassis. They are generated from an annotated genome by expressing the metabolic network of an organism as a list of mass balanced reactions, which is represented as a so-called stoichiometric matrix. They also relate genes with the proteins they code for and the reactions those proteins are associated with. They helped satisfy two main purposes: describe the metabolism - specifically, a steady-state approximation thereof - and predict the state of a metabolic network in different scenarios.
How do we use metabolic models?
There are many mathematical analyses that can be performed on GSMMs; the method we are using is called flux-balance analysis (FBA).
Setting flux constraints
For each reaction, the range of allowable fluxes is decided based on biological data, the simulation target, and heuristics. A steady-state assumption is also made that couples the reaction fluxes based on the topology and stoichiometry of the GSMM network.
Conducting Flux-Balance Analysis
Flux-balance analysis (FBA) assigns each reaction a flux such that (A) the flux constraints are obeyed and (B) a given objective function is optimized (often set to be a predictor of cellular growth). We aim to use FBA and its associated methods for the in silico discovery of genes that can be added, overexpressed, or knocked out to improve methanol-dependent growth.
Insights from Flux-Balance Analysis
FDA generates insights on:
- Knockouts
- Overexpression
- Concentration of Growth Media and Supplements Required
Knockouts
Gene knockout simulations are performed by systematically removing each gene from the model and re-optimizing the flux distribution using FBA. This involves setting the fluxes associated with the reactions catalyzed by the knockout gene to zero or adjusting them to account for the gene's absence. The impact of each gene knockout on the objective function is assessed by comparing the predicted flux distribution and the optimized objective function value with those of the wild-type (unmodified) model. Genes whose knockout results in a significant decrease in the objective function are considered potential targets for further experimental validation.
Overexpression
Overexpression simulations are performed with maximizing the production of a target metabolite in mind. From the dry-lab perspective, we identified potential overexpression targets using a framework called FSEOF, and are trying to simulate overexpression in a constraint-based modeling python package called COBRApy with an additional imported metabolic network design package called StrainDesign [10] . In order to simulate overexpression in our T-B18 model, we plan to follow the procedure outlined in slides 25-27 of this powerpoint , where we obtain the range of flux values (that exist in the solution space) for reactions that we want to overexpress. Afterwards, we set the ‘lower-bound’ attribute of a reaction to be the maximum value (in the range) from the FVA analysis. The upper-bound can be set to some arbitrarily large value (1000+). By doing this, we essentially constrain the model to high fluxes for reactions of interest that we intend to overexpress.
Concentration of Growth Media and Supplements Required
From the simulation perspective, we define the growth medium as the fluxes of exchange reactions between the bacteria and the environment. We can add reactions, limit fluxes, and eliminate reactions to simulate different growth media conditions.