Cattlelyst is built on three pillars: ammonia conversion, methane oxidation, and safety. Each pillar encompasses several wetlab projects. On this page, we present and highlight the design and results of our work in the lab. Detailed reports on the separate wetlab projects can be found on the Wetlab page or via hyperlinks within this page.
The first pillar of Cattlelyst is based on the conversion of ammonia into dinitrogen gas. We found bacteria able to convert ammonia to dinitrogen gas, in a process called heterotrophic nitrification-aerobic denitrification (HNAD). However, these will most likely not survive in our biofilter, because their natural conditions differ a lot from ours. Being synthetic biologists, we decided to design our own bacterium that can consume ammonia and produce dinitrogen. Before we entered the lab, we developed a dynamic metabolic model to increase our conceptual understanding of HNAD. This process is divided in two different processes, namely nitrification and denitrification. During nitrification, ammonia (NH3) is converted into nitrate (NO3-), nitrite (NO2-) and/or nitric oxide (NO) . Denitrification is the reduction of NO3- or NO2- to the gaseous nitrogen compounds nitric oxide (NO), nitrous oxide (N2O) and N2 , . N2O is one of the intermediates in the denitrification pathway and has a global warming potential 300 times that of carbon dioxide over 100 years (Figure 1). We applied this knowledge in the design of our lab experiments by establishing two important goals: 1) engineering the entire nitrification and denitrification pathways in one organism and 2) limiting N2O production. As our chassis, we chose Pseudomonas putida (P. putida), which is a non-denitrifying, non-pathogenic model organism . For more information, see the engineering cycle page.
Although the exact mechanisms or pathway for nitrification are not known, the conversion of NH3 to hydroxylamine (NH2OH) and subsequently to NO2- or NO facilitated by ammonia monooxygenase (AMO) and hydroxylamine oxidoreductase (HAO), respectively , . To date, the genomic sequences of AMO and HAO from heterotrophic nitrifiers have never been fully characterized . Therefore, we used the genes from the well characterized autotrophic nitrifier Nitrosomonas europaea.
Two other iGEM teams have attempted to heterologously express the N. europaea nitrification genes. The Virginia 2017 team tried to express both AMO and HAO with the corresponding cytochromes (c554 and cycX) from N. europaea in Paracoccus denitrificans. However, they only succeeded in expressing HAO properly. The DTU Denmark 2013 team tried as well, but only succeeded in expressing AMO in Escherichia coli (E. coli). Neither team used codon harmonization. Therefore, we decided to apply codon harmonization in our chassis P. putida for both AMO and HAO, and their corresponding cytochromes (c554, c552 and cycX). The different subunits or genes of AMO or HAO were combined on a plasmid into one operon with a synthetic ribosome binding site (RBS) between each subunit. All plasmids were successfully cloned into P. putida (Figure 2).
Because the exact mechanism and product of nitrification are ambiguous, we decided to introduce the complete denitrification pathway in P. putida. It consists of four enzymes: nitrate reductase (Nap), nitrite reductase (Nir), nitric oxide reductase (Nor), and nitrous oxide reductase (Nos). We developed two approaches to reach this goal: the mosaic approach, and the plug-and-play approach.
This approach is similar to the nitrification approach: different genes were combined on a plasmid into one operon with synthetic RBSs between each gene. This way, the enzymes could be separately engineered and tested. Then, active enzymes could be combined in one strain. We aimed to optimize the denitrification pathway and its regulation in a way that NO3- or any other intermediate is funneled through the pathway, avoiding accumulation of (other) intermediates, This should maximize the conversion rate of NH3 to N2.
In contrast to nitrification, we could use the genes from known aerobic denitrifiers, as their denitrification genes are characterized. Enzymes from the following known aerobic denitrifiers were tested: Paracoccus denitrificans and Pseudomonas stutzeri –. Additionally, the genes from the denitrifier Cupriavidus necator were used. This bacterium is not known for aerobic denitrification, but it does possess Nap, which is specifically known for being active in aerobic conditions . By using the genes from different organisms, multiple variants of the enzymes could be tested.
Within the timeframe of the iGEM competition, we were able to successfully express the Nap and Nir enzymes in one strain. This strain was grown with NO3- for 24 hours (Figure 2). Clearly, more NO2- is produced with Nap alone in comparison to Nap + Nir. This indicates that the produced NO2- from Nap is consecutively consumed by Nir. These results indicate that the enzymes are simultaneously functional (Figure 3).
Besides the mosaic approach, we created a bacterial artificial chromosome (BAC), carrying all four denitrification operons from P. stutzeri. To guarantee complete transcription of the 31 kb cargo, a T7 promoter was added in the middle, between the Nir and Nor operons. Moreover, at the 5’ end of each operon an RBS was added. High copy number plasmids are normally employed to continuously propagate the system. However, high copy number plasmids containing a large cargo can excessively burden their host and are prone to mutate . Therefore, we integrated the complete cargo in the genome of P. putida EM42 ∆nasT.
To transport the complete denitrification machinery, we prepared a ‘landing pad’ in P. putida EM42 ∆nasT. The landing pad comprises a T7 promoter that controls the expression of Cre recombinase, flanked by two lox sites. The denitrification machinery and gentamycin resistance cassette on the BAC were flanked by compatible lox sites as well. After successful conjugation of the BAC into P. putida EM42 ∆nasT, transiently expressed Cre recombinase recognized the lox sites and subsequently integrated the denitrification machinery.
Given that we placed the denitrification pathway under the control of 2 T7 promoters, we needed to integrate a T7 polymerase (T7pol). This polymerase specifically transcribes DNA only downstream of a T7 promoter, synthesizes RNA at a high rate. Additionally, the T7pol ignores terminators, which increases the likelihood that the whole pathway is transcribed . Moreover, we made T7-polymerase expression IPTG inducible. This was done to prevent constitutive expression, and with that the associated metabolic burden of the denitrification cassette. The final strain was called P. putida :SD, SD standing for Synthetic Denitrification.
Within the timeframe of the iGEM competition, we were able to test NO2- and N2O accumulation. For NO2- accumulation specifically, the strains were grown with NO3- for 24 hours. Furthermore, we tested the effect of IPTG on nitrogen dynamics. The strains were grown with with IPTG (:SD+) and without IPTG (:SD). For the :SD+ condition, the pathway is transcribed to a higher extend. Compared to the control, more NO2- accumulates for :SD and :SD+, suggesting that the NO3- reductase works. However, when compared with the Nap plasmid, less NO2- accumulates (Figure 4). This could imply two things: (1) nitrate reduction is not as optimal as for the plasmid or (2) the coupled NO2- reductase works too. The fact that the :SD with IPTG accumulates less NO2- compared to the :SD suggests that more NO2- is reduced. This could be explained by higher expression rates downstream of the pathway.
To test whether the NO2- reductase, and nitric oxide reductase work, we performed a gas chromatography-mass spectrometry (GC-MS) experiment during which we measured the N2O in the headspace. The N2O in normal air is about 8 ppm. We discovered that after 139 hours no N2O accumulates for :SD without IPTG. However, for two out of three biological replicates for :SD + IPTG we did see a slight increase in N2O to 11 and 14 ppm. These findings hint that Nap – Nir – Nor work for P. putida :SD + IPTG.
Denitrification: limiting N2O accumulation
For both the mosaic and plug-and-play approach, we designed an additional project based on CRISPR interference (CRISPRi). With CRISPRi we aimed to redirect the electron flux towards the denitrification machinery.
We learned that the distinct denitrification steps influence each other through electron competition. Pan et al.  have shown that the competition arose when the supply of electrons did not meet the demand for electrons by the four reductive denitrification steps. During electron-limiting conditions, electrons are allocated differently to the enzymes, consequently leading to N2O accumulation .
P. putida normally uses oxygen respiration as its only electron sink. However, in P. putida :SD, we added the denitrification pathway which is an additional electron sink. By having two electron sinks, we increase the risk that there are insufficient electrons left to fuel N2O reduction to N2, leading to N2O accumulation. Aerobic respiration in P. putida depends on five terminal oxidases, catalyzing the four-step reduction of oxygen to water. To impair P. putida’s ability to dissipate electrons through oxygen respiration, we developed 15 different CRISPRi plasmids targeting the region between the promoter and start codon of these complexes. Given that these genes are essential, we tested for all 15 plasmids if growth was affected (Figure 5). Spacer 4, targeting the promoter of Cyo oxidase, impaired growth during the exponential phase.
The second pillar of Cattlelyst is based on the consumption of methane gas (Figure 6). We discussed why natural methanotrophs such as Methyloccocus capsulatus are not an option in the Engineering cycle. As such, we decided to build a synthetic methanotroph in E. coli, However, before entering the lab we made sure to computationally verify the methane conversion pathway in E. coli using the iGEM PIPE. We used this knowledge to design the pathway for synthetic methanotrophy in C1 consuming strains, that either grown on methane derived products, methanol, formaldehyde, or formate. Both literature and the iGEM PIPE suggested using a methane monooxygenase (MMO) of which two variants exist: the particulate and soluble MMO. The difference between these variants is extensively discussed in the MMO Wetlab page and Engineering cycle page. For our project we followed a multifaceted approach, working to express both variants in E. Coli strains that can use C1 compounds. This would result in a synthetic methanotroph able to consume methane and turn it into biomass or carbon neutral carbon dioxide, an important aspect within the Cattlelyst biofilter.
C1 growing strains
The conversion of methane to methanol does not complete a synthetic methanotroph: methanol has to be converted into either biomass or carbon dioxide. Here two strains that convert methanol or formaldehyde are used. The first strain, C1Saux, is auxotrophic for formaldehyde if grown on minimal medium as it requires formaldehyde or formateto make glycine via the reductive glycine pathway. Additionally, this strain requires the enzyme methanol dehydrogenase (Mdh) to convert methanol into formaldehyde, which has been expressed as shown in . After transformation of this C1Saux strain with the Mdh plasmid, it was able to grow on minimal medium with methanol, as shown in the wetlab MMO page. The second strain is the SM1 strain, which exhibits the Ribulose monophospate pathway and as such can grow with methanol as sole carbon source. Adding a MMO that works in vivo to these strains would allow them to grow on methane. Unfortunately, due to time restraints there was no opportunity to test this.
Two types of MMO have been engineered within this project. First for the sMMO variant, a similar approach to both the E. Cowli strain from iGEM Braunschweig 2014 and the Bennet et al. 2021 paper is used . Here the entire operon from the M. capsulatus species is taken and expressed together with chaperone proteins. Unfortunately, this aspect never got out of the cloning phase and thus this project was never tested. The second approach is the heterologous expression of the pMMO, for which the method of Kim et al. 2019 was used . This proved more fruitful as the simpler cloning steps allowed us to test this enzyme both in vivo and in vitro. Using Sonification, filtration and His-Tag purification of the product the presence of the enzyme was suggested on SDS-page as can be seen in Figure 7. No enzyme activity was detected after 25 hours of incubation of the purified enzyme in a methane rich atmosphere. After iGEM this project will continue and other mimics of pMMO will be expressed and tested in in vitro . Additionally, sMMO will be expressed in one of the C1 growing strain, engineering a synthetic methanotroph.
Because we designed our biofilter Cattlelyst for a real-world application on cattle farms, biocontainment of the GMOs is crucial. Therefore, we designed three layers of safety. The first layer is based on the higher methane concentration inside the biofilter than outside of it. This concentration difference is coupled to a methane-dependent kill switch. Another characteristic of the biofilter is the high cell density inside, so the second layer of safety is a proximity kill switch, which is linked to the methane-based kill switch. Thirdly, a co-dependency is added, which makes E. coli and P. putida dependent on each other. As a result, they would starve if they escape the biofilter by themselves. The combination of these three layers of safety ensures total dependency of the two bacterial species on the biofilter’s conditions and on each other, making them unable to escape from the biofilter.
Methane-dependent kill switch
The first layer of our safety mechanisms creates a conditional kill switch in the synthetic methanotroph E. coli that relies on the methane concentration present, which is higher inside the biofilter than outside. A central intermediate of the methane conversion pathway, formaldehyde (CH2O), is used to base the system on , . As formaldehyde is a highly (geno)toxic compound , it is rapidly detoxified in wild-type E. coli by the FrmRAB operon, consisting of the detoxification genes frmA and frmB, controlled by the repressor frmR of its corresponding promoter Pfrm . This system provides a sensitive formaldehyde biosensor , which is coupled to the Hok-Sok toxin/antitoxin system. The production of the toxin Hok can kill the cell by disrupting its membrane structure. It is regulated by a short-lived antisense RNA, Sok, which prevents translation of the hok mRNA .
A genetic circuit is designed that contains the FrmR protein as formaldehyde sensor, which is coupled by means of LacI to the Hok/Sok system. This ensures toxin production and cell death in low methane concentration, and cell survival in high methane concentrations, functioning as a conditional ‘kill switch’ in E. coli. Visit the methane-dependent kill switch page or a more detailed description.
To ensure the biofilter bacteria do not die when the methane concentration in the biofilter temporarily drops, we couple the production of toxin to our second safety mechanism based on cell density (see below) (see our human practices work om how we came to this idea). This is achieved by means of a hybrid promoter, that responds to input signals from both LacI (methane-dependent kill switch) and LuxR (proximity-based kill switch). This way, the bacteria only die when both methane and cell density are low.
This system was designed as a plasmid to be constructed in E. coli, but due to cloning difficulties and time constraints, the complete plasmid was not obtained and its dynamics could not be tested. Additionally, to enable increasing the formaldehyde sensitivity of the system, toxicity tests are performed on several strains containing knock-outs of detoxification genes.
Proximity-dependent kill switch
The second of our safety mechanisms is the proximity-dependent kill switch. The cell density can be ‘sensed’ by using quorum sensing. Quorum sensing is the ability of microorganisms to regulate gene expression upon different cell density by exchanging signaling molecules . In Vibrio fischeri the quorum-sensing molecule (AHL) is produced by the LuxI (BBa_C0061) protein. This molecule binds to LuxR, which can either activate the lux pR promoter (BBa_R0062) or repress the lux pL promoter (BBa_R0063) , . In our envisioned circuit, a toxin is put under control of the lux pL promoter and the corresponding antitoxin under control of the lux pR promoter as explained on the proximity based kill switch page. We have shown that AHL rich medium is able to both activate the lux pR promoter and repress the lux pL promoter when they are put together in our reporter strain (see Figure 17). These results show that this safety circuit could ensure biocontainment of the microorganisms as it makes them dependent on the high cell density in the biofilter.
For the functioning of the biosafety circuit, the toxin/antitoxin ratio in high cell densities versus low cell densities is important . To obtain a sufficient big change in the toxin/antitoxin ratio in high versus low cell densities, the circuit needs to be sensitive to an input signal, in this case AHL concentration. The sensitivity of the circuit to the AHL concentration (with is related to the cell density) is calculated by:
As can be seen in Figure 9, the circuit is sensitive to the cell density and the sensitivity increases over time.
A co-culture experiment was performed to investigate if P. putida can activate the quorum sensing circuit in E. coli. An AHL producing P. putida strain and a wild type P. putida strain as negative control were spread over agar plates and drops of different dilutions of the reporter E. coli strain were added on top. Pictures of the fluorescence of the plates showed activation of the reporter E. coli strain when grown together with a AHL producing P. putida strain.
Co-dependency was engineered in addition to the methane dependent kill switch and the proximity kill switch to add another layer of safety to the Cattlelyst biofilter. Co-dependency of the two bacterial species was based on the establishment of a cross-feeding community of Escherichia coli and P. putida reliant on amino acids exchange and carbon-source dependency.
Amino acids exchange
Following a preliminary study two amino acids were selected for the establishment of the cross-feeding community: tryptophan (Trp) and tyrosine (Try). Specifically,
P. putida was chosen to be auxotrophic for Tyr and overproducing Trp, while E. coli was selected to be auxotrophic for Tyr and overproducing Trp.
Overall, the single mutant strains P. putida EM42 ΔtyrA (partial auxotroph for Tyr), P. putida EM42 ΔphhAB (partial auxotroph for Tyr), E. coli ΔtrpD (auxotroph for Trp) and E. coli ΔtyrR (overproducer of Tyr) were obtained. Nevertheless, due to time constrains, the full cross-feeder strains could not be created. Anyway, via a biosensor experiment, overproduction was studied in the single mutants P. putida EM42 ΔtyrA and E. coli ΔtyrR. P. putida EM42 ΔtyrA and E. coli ΔtyrR were found to overproduce Trp and Tyr respectively. Overproduction of Trp and Tyr was detected to be in high enough quantities to support the growth of the E. coli auxotrophic strains used as biosensors in the experiment (see figure 11). These preliminary results indicate that both P. putida EM42 ΔtyrA and E. coli ΔtyrR could sustain the minimal growth of the cross-feeding partner.
Carbon-source dependency between E. coli and P. putida
Carbon-source dependency was included to confer an additional layer of safety to the system. In this context, P. putida was engineered to rely solely on E. coli for the carbon-source necessary to accumulate biomass. E. coli was shown to excrete acetate both during anaerobic and aerobic growth, and P. putida is known to grow on acetate as sole carbon-source. Thus, a P. putida strain solely reliant on acetate for biomass generation was used in this project. This strain presents a double knock-out for the operon gtsABCD (encoding the ATP-dependent ABC glucose transporter) and for the gene gcd (encoding the glucose dehydrogenase. The two knock-out are reported to impede the uptake of glucose by preventing its transport through the inner membrane mediated by the ABC transporter and by blocking the peripheral oxidative route, which, through the action of Gcd, converts glucose to the intermediate D-gluconate.
Most importantly, co-culture experiments showed that P. putida EM42 significantly increased in abundance after 24 hours of co-culture with E. coli (Figure 12). Conversely, the co-culture of P. putida EM42 Δgts Δgcd and E. coli was shown not to significantly differ from the initial inoculation ratio after 24 hours of growth (Figure 12). P. putida EM42 Δgts Δgcd was postulated to maintain a lower frequency in the population due to its dependence on the acetate secreted by E. coli.
Overall, carbon-source dependency was shown not only to serve as an ulterior layer of safety, but to successfully maintain the initial inoculation ratio of the two bacterial species after 24 hours of co-culture.
- C. JD and L. KM, “Nitric oxide is an obligate bacterial nitrification intermediate produced by hydroxylamine oxidoreductase,” Proc. Natl. Acad. Sci. U. S. A., vol. 114, no. 31, pp. 8217–8222, Aug. 2017.
- G. Giannopoulos et al., “Tuning the modular Paracoccus denitrificans respirome to adapt from aerobic respiration to anaerobic denitrification,” Environ. Microbiol., vol. 19, no. 12, pp. 4953–4964, 2017.
- N. Shearer, A. P. Hinsley, R. J. M. Van Spanning, and S. Spiro, “Anaerobic growth of Paracoccus denitrificans requires cobalamin: Characterization of cobK and cobJ genes,” J. Bacteriol., vol. 181, no. 22, pp. 6907–6913, 1999.
- Z. Tan, Air Pollution and Greenhouse Gases. Springer, 2014.
- M. Martin-Pascual et al., “A navigation guide of synthetic biology tools for Pseudomonas putida,” 2021.
- J. M. Wehrfritz, A. Reilly, S. Spiro, and D. J. Richardson, “Purification of hydroxylamine oxidase from Thiosphaera pantotropha. Identification of electron acceptors that couple heterotrophic nitrification to aerobic denitrification,” FEBS Lett., vol. 335, no. 2, pp. 246–250, 1993.
- and K. M. L. Caranto, Jonathan D., “Nitric Oxide Is an Obligate Bacterial Nitrification Intermediate Produced by Hydroxylamine Oxidoreductase,” Proc. Natl. Acad. Sci. U. S. A., vol. 114, no. 31, pp. 8217–22, 2017.
- D. E. Liu, X., Shu, Z., Sun, D., Dang, Y., & Holmes, “Heterotrophic nitrifiers dominate reactors treating incineration leachate with high free ammonia concentrations,” ACS Sustain. Chem. Eng., vol. 6, no. 11, pp. 15040–15049, 2018.
- J. Guo et al., “Pathways and organisms involved in ammonia oxidation and nitrous oxide emission,” Crit. Rev. Environ. Sci. Technol., vol. 43, no. 21, pp. 2213–2296, 2013.
- Z. Zhu et al., “Quorum sensing systems regulate heterotrophic nitrification-aerobic denitrification by changing the activity of nitrogen-cycling enzymes,” Environ. Sci. Ecotechnology, vol. 2, p. 100026, Apr. 2020.
- M. Miyahara et al., “Potential of aerobic denitrification by pseudomonas stutzeri TR2 to reduce nitrous oxide emissions from wastewater treatment plants,” Appl. Environ. Microbiol., vol. 76, no. 14, pp. 4619–4625, Jul. 2010.
- K. Medhi, A. Singhal, D. Chauhan, I. T.-B. technology, and undefined 2017, “Investigating the nitrification and denitrification kinetics under aerobic and anaerobic conditions by Paracoccus denitrificans ISTOD1,” Elsevier.
- T. S, “Expression using the T7 RNA polymerase/promoter system,” Curr. Protoc. Mol. Biol., vol. Chapter 16, no. 1, Jul. 2001.
- Y. Pan, B.-J. Ni, P. L. Bond, L. Ye, and Z. Yuan, “Electron competition among nitrogen oxides reduction during methanol-utilizing denitrification in wastewater treatment,” 2013.
- S. Wenk et al., “An ‘energy‐auxotroph’ Escherichia coli provides an in vivo platform for assessing NADH regeneration systems,” Biotechnol. Bioeng., vol. 117, no. 11, pp. 3422–3434, Nov. 2020.
- R. K. Bennett et al., “Expression of soluble methane monooxygenase in Escherichia coli enables methane conversion,” bioRxiv, p. 2021.08.05.455234, Aug. 2021.
- H. J. Kim et al., “Biological conversion of methane to methanol through genetic reassembly of native catalytic domains,” Nat. Catal. 2019 24, vol. 2, no. 4, pp. 342–353, Apr. 2019.
- B. W. Hütsch, “Tillage and land use effects on methane oxidation rates and their vertical profiles in soil,” Biol. Fertil. Soils, vol. 27, no. 3, pp. 284–292, Jul. 1998.
- F. Y. H. Chen, H. W. Jung, C. Y. Tsuei, and J. C. Liao, “Converting Escherichia coli to a Synthetic Methylotroph Growing Solely on Methanol,” Cell, vol. 182, no. 4, pp. 933-946.e14, Aug. 2020.
- H. Yurimoto, N. Kato, and Y. Sakai, “Assimilation, dissimilation, and detoxification of formaldehyde, a central metabolic intermediate of methylotrophic metabolism,” Chem. Rec., vol. 5, no. 6, pp. 367–375, Jan. 2005.
- K. J. Denby et al., “The mechanism of a formaldehyde-sensing transcriptional regulator,” Sci. Rep., vol. 6, no. 1, pp. 1–15, Dec. 2016.
- B. M. Woolston, T. Roth, I. Kohale, D. R. Liu, and G. Stephanopoulos, “Development of a formaldehyde biosensor with application to synthetic methylotrophy,” Biotechnol. Bioeng., vol. 115, no. 1, pp. 206–215, Jan. 2018.
- K. Gerdes, L. K. Poulsen, A. K. Nielsen Novozymes, and J. Martinussen, “The hok killer gene family in gram-negative bacteria Regulation of stringent response in E.coli View project FAST and BM4SIT and the ESA-ITN project View project,” 1990.
- M. B. Miller and B. L. Bassler, “Quorum Sensing in Bacteria,” Annu. Rev. Microbiol., vol. 5, no. 1, pp. 165–199, 2001.
- W. R. J. D. Galloway, J. T. Hodgkinson, S. D. Bowden, M. Welch, and D. R. Spring, “Quorum sensing in Gram-negative bacteria: Small-molecule modulation of AHL and AI-2 quorum sensing pathways,” Chem. Rev., vol. 111, no. 1, pp. 28–67, 2011.
- J. Zhang et al., “Binding site profiles and N-Terminal minor groove interactions of the master quorum-sensing regulator LuxR enable flexible control of gene activation and repression,” Nucleic Acids Res., vol. 49, no. 6, pp. 3274–3293, 2021.
- J. Peng and L. R. Triplett, “Activation of metabolic and stress responses during subtoxic expression of the type I toxin hok in Erwinia amylovora,” pp. 1–22, 2020.