Metabolic Engineering

Synthetic Metabolism: Pathway Design

The ultimate goal of synthetic metabolic engineering is to understand a cellular chassis metabolism so well that we can design a complex pathway that doesn’t exist in nature for the production of a novel product or fuel that shows 100% efficiency within its heterologous host. In addition, these complex pathways could be catalyzed with enzymes that also do not exist in nature.

However, this is a tall order. Natural systems add layers of complexity to chemical reactions through evolutionary innovation, such as regulatory mechanisms, protein colocalization, kinetics, feedback loops. Therefore, when designing metabolic engineering projects, it is often best to use the optimized parts from a relevant natural pathway as much as possible. It follows that the most successful ventures as of yet are the enhancement of natural pathways to make the desired product more efficiently. This can be undergone by metabolic bypassing (skipping a limiting step), replacement (replacing a limiting step with a more efficient enzyme), or orthogonality (almost completely replacing all of the natural steps).

Pathway Design

Today, the KEGG database collects information about all known metabolic pathways and enzymes to aid synthetic pathway design. There are different levels of pathway complexity one can design. A pathway can be designed by replicating a similar pathway design already found in nature, or by combining different parts of natural pathways together. The more elaborate the pathway, the more likely we are to use a natural pathway as found in its endogenous host, as the less likely it is that we can understand all the regulatory components to reproduce them somewhere else. The simpler the pathway to a product, the more likely we can design the entire pathway from scratch, even if it does not yet exist in nature.

The easiest approach to take in synthetic metabolism is the copy and paste approach, which involves transferring existing pathways (enzymes) into a new host. This is especially useful for making high-value products that are otherwise sourced from unsustainable/limiting supplies. For example, vanilla is made from fossil fuels, but its flavor can be made with the enzyme vanillin aldehyde transferred into yeast from vanilla pod plants (Hansen et al. 2009). However, the copy and paste approach is limited to the structure of naturally existing pathways. A mix and match enzyme approach allows the creation of a new pathway, by recombining the repertoire of existing enzymes to form new synthetic pathways that redirect flux to make new products. For example, the Jones lab at Imperial College London (Kallio et al. 2014) used the mix and match with enzymes from different species to create a microbe to produce the first biological source of renewable propane. Propane is the main component of liquid petroleum, and so propane biofuel could be easily incorporated into existing infrastructure for transportation, storage, and utilization to replace unrenewable propane.

Optimising Catalysts

The use of novel reactions/novel enzymes is the most difficult to carry out and a future challenge of metabolic engineering. Such pathway design has the potential to increase the space of possible metabolic solutions dramatically. Rational engineering of enzymes is a growing procedure in research to optimize efficiencies and alter substrate specificity. If you start with a known enzyme that catalyzes your pathway of interest, directed mutagenesis rounds will allow its rate optimization. Mutagenesis is directed at the sites that interact with the substrate, which often requires an existing structure resolution of the enzyme. If no known enzyme catalyzes your pathway of interest, it requires choosing a similar enzyme (in the same class, with a similar substrate), and screening for promiscuous activity after mutagenesis rounds. These candidates can be found in previous literature searches, databases (for example, BRENDA focuses on enzymes), and phylogenetic analyses.

An example of the power of rational engineering in synthetic metabolism is the work of Schwander et al 2016. This group changed the substrate of an enzyme so that it could be added to complete the cycle of a synthetic pathway designed in vitro for CO2 fixation. After several rounds of mutagenesis, their screen found a variant that accepted their desired substrate, and they subsequently optimized this variant’s catalysis with further mutagenesis. This work created the most complex synthetic pathway to date, containing 17 enzymes from all 3 domains of life, but is yet to be accomplished within a microbial host (natural metabolic systems are a complex background for synthetic pathways…). If this carbon fixating pathway can be transferred to a microbial host, it will aid increased rates of carbon fixation.

The design of novel enzymes from scratch is still only an idea but would breach the limits that nature imposes on all the approaches to synthetic pathway design above. The use of computational tools and software enables ab initio design and evaluation of a new pathway, with new or existing enzymes, for example, ATLAS, RetroPath, and FMM. These tools allow host selection and can propose the use of native pathways as well as synthetic ones. Finding the best design of a pathway of interest requires comprehensive searching through many databases, and the use of many tools, as they are new and constantly updating as the field of metabolic engineering expands.

Microbial Hosts for Synthetic Metabolism

Appropriate chasses

A chassis cell provides a physiological, metabolic, and regulatory environment in which one can switch in and out synthetic pathways and metabolic networks. Each chassis type has a different complex systems macro network within which the synthetic pathway or circuit must fit. Only a few cellular chasses have been used to achieve commercial exploitation by engineering them for specific product outputs. This is because there have been issues with core metabolic engineering, yields, titers, and productivity, that often make a process unfeasible when compared to a non-renewable method of obtaining the same product.

Industrial metabolic engineering applications often involve complex engineered properties, requiring whole sets of different gene additions that have roles in multiple biological processes. A common design scenario for industrial application is that all enzymes or pathways required for the complex property of interest already exist in nature, but no one natural organism contains all components. Therefore an organism containing part of the property potential must be engineered with the rest of the function taken from another organism. In principle, the cell chassis chosen out of the two must be the easiest engineering challenge. However, there are many properties of an ideal chassis that go beyond model organisms’ advantages of genomic manipulation ease. The chassis must support the activity of the engineered exogenous genetic component without interfering with its ability to grow. In addition, ideal chasses generate robust cellular envelopes for bioreactor conditions, have well-defined metabolic flux networks, and have native secretion systems that can secrete the desired metabolite for efficient harvesting.

An absence of evolutionary processes that could affect the performance of the exogenous circuits is also key. This includes avoiding the effects of metabolic burden and catabolic repression that could occur when adding a synthetic pathway to a chassis. To overcome these barriers, linking the exogenous DNA to the cell host’s ability to survive, ensures the transgenes will not be repressed from expression by fierce competition with the host’s transcription machinery. This can be undergone by identifying trade-offs that occur naturally in the host and linking the expression of exogenous DNA with the trait (a bit like the coupling of cofactor balance with a pathway of interest discussed in the ‘cofactor balance’ section).

All of these factors must be taken into account in order to choose a chassis that will allow the economic and technical feasibility of a metabolic engineering venture. Therefore, though E. coli, Saccharomyces cerevisiae, and Bacillus subtilis are well-understood model organisms used in academic synthetic biology and metabolic engineering studies, with large repertoires of molecular tools available for their engineering, most industrial biotechnology does not use them as chasses. Examples of non-model organisms being utilised for biotechnological applications are discussed in the ‘bioremediation and biorefining’ page of this blog. If the metabolic engineering field continues to expand, it will require the identification and genomic and metabolic characterization of new species found in nature to use as new, suitable chasses.

Genetic engineering of non-model organisms

The transfer of an exogenous DNA circuit into a non-model organism requires species specific molecular tools for precise genomic modifications, physical transformation, and maintenance of the exogenous DNA on a plasmid in the host or stable maintenance of chromosomally integrated DNA. Physical transformation techniques for bacteria include natural competence (growing certain microbes under specific conditions in which they uptake free floating DNA from the environment), electroporation (especially good for gram negative bacteria, DNA uptake by inducing cellular electric shock) and conjugation (using conjugative plasmid templates to contain the DNA of interest for delivery). Often, the building of custom electroporation equipment for a specific non-model organism is necessary.

For most non-model organisms, the development of tools for the maintenance and manipulation of exogenous and endogenous elements is ongoing. Systematic design of these tools ensures that when complete they will greatly expand the diversity and amount of biotechnological applications using them because of their ease of use to transfer amongst laboratories. The most obvious example of systematic design is that of bacterial plasmids, which have been used since the very beginning of molecular biology as tools for interspecies transfer of DNA. Since then, conventional plasmids have been refactored to become modular, each module flanked by standardised restriction endonuclease sites. This allows mass use of the same plasmids backbone, but a mix and match strategy of the components within it for particular goals. For example, the Standard European vector architecture (SEVA) plasmid contains a cargo block (to clone in genes of interest ), an oriT (to make the plasmid transferable by conjugation), an ori (for replication, which can be changed for plasmid compatibility modifications), and an antibiotic resistance selectable marker.

SEVA (Standard European Vector Architecture) plasmid

This standardisation of plasmid architecture is a good example of how systematic design of molecular tools allow for their mass use, to save time and money as we enter the systems and synthetic biology era. It also allows for species transferability in the context of industrial biotechnology, when applying E. coli built synthetic gene circuits into more appropriate factory chassis hosts that are less well characterised. To use standard architecture in different cellular systems, it is necessary to determine its transfer frequencies, characterise the different antibiotic concentrations each transformed species can tolerate, and its segregationist stability, which can all change on host transfer. Segregation stability is important; some strain populations may retain the plasmid for many generations (for example, 99% rate of daughter cell transference), or it may decay quickly through time (67%). Plasmids are considered stable within a species if they stay in the population for more than 100 generations, and plasmid stability can be increased by applying antibiotic selection pressure to the population if necessary.

An example of plasmid design for use in a specific non-model chassis was from Heap et al. 2009, who designed a standardized modular system for a Clostridium-Escherichia coli shuttle plasmid, containing a novel clostridium replicon. Clostridium species are more relevant for medical and industrial application than E. coli, and this vector can now be used by the whole research community for academic and industrial applications to transfer engineered properties from E. coli into Clostridium.

A big challenge for exogenous gene transfer into host cells is the restriction barrier. This refers to a species use of restriction enzymes as a form of immune response, that cut up foreign invading DNA (often phages in nature), that can target plasmids introduced in the lab. Bacterial methylases methylate endogenous DNA to prevent it from also being recognised by restriction enzymes. Many approaches have been suggested for overcoming the restriction barrier. Firstly, incorporating a methylase that methylases the plasmid DNA for digestion protection can be undergone. Mermelstein and Papoutsakis (1993) demonstrated the effectiveness of this approach, showing the phi 3T I methyltransferase can protect three different plasmids against Cac824I restriction in clostridium after electroporation. As methylated sites are dependent on the methylase type, and digestion sites are dependent on the restriction enzyme employed by the bacteria, the restriction enzyme type must be known for this approach to work. Otherwise, methylation may not hide digestion sites on the transgenic plasmid. In addition, a methylation step in vitro is necessary before plasmid transformation into the cell, which is not always compatible with the delivery system.

A second approach is to removing restriction sites altogether when designing a plasmid. This approach requires the identification of the restriction enzymes and their sites employed in the host bacteria. Its an especially useful strategy if no protective methylase for the restriction enzyme site has been discovered. Replacing plasmid segments with non-restriction enzyme sites used to be very difficult, but is becoming easier with DNA synthesis and assembly technological advances. Silencing restriction sites on the plasmid of interest, instead of removing them is also a possibility. This involves utilizing the redundancy in the genetic code; changing the DNA sequence at the third position of each codon so that restriction enzymes can no longer digest it, but without changing the same amino acid sequence.

If inclusion of a restriction site in the plasmid is a necessity (for example if it is in a promoter region), an alternative approach is to knock out the restriction enzyme systems in the bacterial host. Dong et al. 2010 engineered a model clostridium species in which its Cac824I restriction enzyme was knocked out by intron insertion, to knock out its activity to digest unmethylated, introduced DNA. Methylase knock out strains of many kinds of bacteria are now available for use in transformation studies. However, knock out strains are therefore vulnerable to phage attack.

References

• Young, J. D., Shastri, A. A., Stephanopoulos, G., & Morgan, J. A. (2011). Mapping photoautotrophic metabolism with isotopically nonstationary 13C flux analysis. Metabolic engineering, 13(6), 656–665.
• Savakis, P. E., Angermayr, S. A., & Hellingwerf, K. J. (2013). Synthesis of 2, 3-butanediol by Synechocystis sp. PCC6803 via heterologous expression of a catabolic pathway from lactic acid-and enterobacteria. Metabolic engineering, 20, 121–130.
• Shen, C. R., Lan, E. I., Dekishima, Y., Baez, A., Cho, K. M., & Liao, J. C. (2011). Driving forces enable high-titer anaerobic 1-butanol synthesis in Escherichia coli. Applied and environmental microbiology, 77(9), 2905–2915.
• Erb, Tobias J et al. “Synthetic metabolism: metabolic engineering meets enzyme design.” Current opinion in chemical biology vol. 37 (2017): 56–62. doi:10.1016/j.cbpa.2016.12.023 (an overview of synthetic metabolism).
• Schwander, T., Schada von Borzyskowski, L., Burgener, S., Cortina, N. S., & Erb, T. J. (2016). A synthetic pathway for the fixation of carbon dioxide in vitro. Science, 354(6314), 900–904.
• Hansen, E. H., Møller, B. L., Kock, G. R., Bünner, C. M., Kristensen, C., Jensen, O. R., … & Hansen, J. (2009). De novo biosynthesis of vanillin in fission yeast (Schizosaccharomyces pombe) and baker’s yeast (Saccharomyces cerevisiae). Applied and environmental microbiology, 75(9), 2765–2774.
• Kallio, Pauli, András Pásztor, Kati Thiel, M. Kalim Akhtar, and Patrik R. Jones. “An engineered pathway for the biosynthesis of renewable propane.” Nature communications 5, no. 1 (2014): 1–8.
• Heap, John T., Oliver J. Pennington, Stephen T. Cartman, and Nigel P. Minton. “A modular system for Clostridium shuttle plasmids.” Journal of microbiological methods 78, no. 1 (2009): 79–85.
• Mermelstein, L. D., & Papoutsakis, E. (1993). In vivo methylation in Escherichia coli by the Bacillus subtilis phage phi 3T I methyltransferase to protect plasmids from restriction upon transformation of Clostridium acetobutylicum ATCC 824. Applied and environmental microbiology, 59(4), 1077–1081.
• Dong, Hongjun, Yanping Zhang, Zongjie Dai, and Yin Li. “Engineering Clostridium strain to accept unmethylated DNA.” PLoS One 5, no. 2 (2010): e9038.