Project Description

Why do we not unlock the full potential of monoterpenoids?

The world is currently facing substantial global issues, with many expected to persist in the years to come. Whether it is food shortage, infectious diseases or the climate crises, humanity needs to find sustainable solutions as soon as possible to enable a safe and healthy life for all humans. Despite the enormous size of the challenges, small organic molecules can offer solutions to those problems.

One particular class with promising applications are monoterpenoids (Fig. 1). Limonene is a potent biopesticide (Liu, Li and Song, 2022), thymol has an anti-bacterial effect on multiple pathogens (Du et al., 2015), and linalool is a jet fuel precursor which can make air travel more sustainable (Mendez-Perez et al., 2017). But large-scale deployment of monoterpenoids is rare. For our iGEM project this year, we wondered about the reasons that monoterpenoids – despite their huge potential – are not applied more often.

The primary source of monoterpenoids are plants, where they play an important role in many physiological processes. Plants’ essential oils contain more than 80% monoterpenoids, but the extraction of essential oils from plants has low yields of ~1% only. In addition, several hundred different compounds can be found in essential oils, making the purification of single monoterpenoids cost-extensive, laborious and inefficient (Asbahani et al., 2015). By conducting interviews with experts in monoterpenoid production and application, we further learned that the composition of monoterpenoids in essential oils heavily relies on the plants’ growth conditions and decrease during plant or oil storage due to their volatility. Therefore, we concluded that monoterpenoid production based on plants is not sufficient to meet current demands in terms of scale and reliability.

Alternatively, we considered chemical synthesis as a more reliable option to unleash large-scale applications of monoterpenoids. Chemical synthesis of high-valuable monoterpenoids is possible (Semikolenov, Ilyna and Simakova, 2001; Amandi et al., 2005), but as for many other natural products, the complex structure of monoterpenoids makes their chemical synthesis difficult and costly (Gao et al., 2020). The laborious chemical synthesis also creates environmental concerns, as many different reaction steps require high energy, polluting solvents and catalysts, and depend on non-renewable resources, like fossil oils (Zhu et al., 2021).

Besides the extraction of monoterpenoids from plants or their chemical synthesis, advances in synthetic biology opened up the possibility of producing monoterpenoids with microorganisms. Bacterial chassis are often the first choice for the heterologous production of fine chemicals. For the production of monoterpenoids, however, working with bacteria has several disadvantages. First, there is a deficiency of the precursor geranyl diphosphate (GPP) in the bacterial host via the native 1-deoxy-D-xylulose 5-phosphate (DXP) pathway, leading to inefficient production of the desired monoterpenoid. To increase the synthesis, the whole mevalonate pathway needs to be transformed into bacteria (Carter, Peters and Croteau, 2003). Second, downstream steps of the monoterpenoid biosynthesis require the functional expression of CYPs, which is challenging in bacterial chassis (Lundemo and Woodley, 2015). Third, even if all heterologous genes can be functionally expressed, monoterpenoid production with bacteria might not be efficient. The accumulation of precursors leads to reduced cell growth, and various monoterpenoids have toxic effects on bacteria by interfering with the cell membrane (Ignea et al., 2019).

Therefore, yeasts, as eukaryotic production strains, have received attention as attractive chassis to produce monoterpenoids. Yeasts have an endogenous mevalonate (MVA) pathway that can be utilized to synthesize monoterpenoids (Ignea et al., 2019). However, the native MVA pathway does not allow efficient monoterpenoid synthesis, as GPP, the precursor of all monoterpenoids, gets rapidly converted to farnesyl diphosphate (FPP). The successive formation of both GPP and FPP in yeasts is catalyzed by the bifunctional enzyme ScERG20. Another limitation when producing monoterpenoids with yeast is product toxicity. Even though yeasts are generally more robust than bacteria, toxic effects of monoterpenoids can limit the possible yields (Zhao et al., 2016).

We, therefore, concluded that there is currently no feasible strategy to produce a wide variety of monoterpenoids at a large scale efficiently. All current methods have limitations, but utilizing yeasts to produce monoterpenoids appeared to be our most promising starting point.

MonChassis - Monoterpenoid production made easy

We developed MonChassis, an adaptable yeast-based platform for large-scale monoterpenoid production coupled with cell-free monooxygenase catalyzation. MonChassis overcomes the current limitations of monoterpenoid production with yeast by metabolic engineering and outsourcing the CYP-catalyzed oxidation to high-valuable monoterpenoids to an electrically-driven cell-free system (Fig. 3).

Concerning the metabolic engineering of yeast, we found that there are many promising approaches to increase the capability of yeast to produce monoterpenoids (Gao et al., 2020). Thus, we decided to build and test the most promising strategies from published literature in the model organism Saccharomyces cerevisiae and the non-conventional yeast Yarrowia lipolytica to create the ideal chassis organism for monoterpenoid production (Fig. 4). Our approaches all shared the strong expression of the rate-limiting steps in the MVA pathway. We overexpressed the two genes ERG13 and a truncated variant of 3-hydroxy-3-methylglutaryl-coenzyme A (tHMGR), which are the bottlenecks in this reaction cascade (Bröker et al., 2018). This increases the flux towards monoterpenoid precursors.

With regard to monoterpenoid precursors, we designed two different metabolic routes for the formation of two different substrates of monoterpene synthases, GPP and neryl diphosphate (NPP) (Fig. 4). GPP is the product of yeast’s MVA pathway and the native substrate for many monoterpene synthases (Degenhardt, Köllner and Gershenzon, 2009). Monoterpene synthases are enzymes of the secondary metabolism, which makes them, on average, less efficient than ScERG20 (Bar-Even et al., 2011). This redirects flux from the monoterpenoid synthesis towards the synthesis of FPP and ergosterols. Thus, for the designed metabolic route utilizing GPP, it is necessary to knockout the ERG20 gene and replace it with enzymes with higher activity for GPP synthesis. We inserted the genes AgtGPPS2 and GgmFPS144 that both display high selectivity for GPP instead of FPP synthesis (Stanley Fernandez, Kellogg and Poulter, 2000; Burke and Croteau, 2002). As FPP formation is vital for yeast, insertion of GgmFPS144 is beneficial as the encoded enzymes still produce sufficient amounts of FPP for cell growth.

The second metabolic route we designed utilizes NPP as the substrate for monoterpene synthases instead of GPP. NPP is a structural isomer of GPP, which many monoterpene synthases can convert, but not Sc ScERG20 (Ignea et al., 2019). By introducing the gene SltNPPS1 from tomato, we created an orthogonal pathway to produce monoterpenoids. However, as SltNPPS1 and ScERG20 compete for the same substrates, it is necessary to decrease the flux towards FPP by downregulating ScERG20. We aimed to achieve this by replacing the native promotors of ScERG20 in the genomes of S. cerevisiae and Y. lipolytica with the copper-repressible promotors pCTR3 and pCTR1, respectively.

While designing the two different metabolic pathways that utilize different precursor molecules for monoterpene synthases, we recognized that the shared limitation of both designs is the need to redirect flux from ergosterol synthesis to monoterpenoid production. As ergosterol synthesis is necessary for cell growth, designs building upon the yeast’s endogenous MVA pathway will always be limited in their yields. Thus, we decided to design a metabolic pathway that decouples monoterpenoid production and ergosterol synthesis. For this, we took advantage of one of the major benefits when working with eukaryotic chassis: Compartmentalization. We designed pathways that introduce heterologous copies of the MVA pathway, that include our modifications of the cytosolic MVA pathway, into the peroxisomes of S. cerevisiae and Y. lipolytica. We chose peroxisomes as our functionalized compartments as they provide favorable physiological conditions to promote the production of monoterpenoid precursors (Dusséaux et al., 2020). In addition to elevated acetyl-CoA levels, the substrate of the MVA pathway, reduced metabolic cross-talk, and the spatial proximity of enzymes increases reaction efficiency (Farhi et al., 2011). Targeting the enzymes to the peroxisomes can be easily achieved by adding the C-terminal peroxisomal targeting signal-1 (PTS1) tripeptide (Brocard and Hartig, 2006). In total, we designed eight different yeast strains (2 precursors X 2 compartments X 2 yeasts) to find the best combination of parameters for MonChassis’s production organism . To this end, we focused on the limited precursor availability for monoterpenoid synthesis in yeast while adressing the problem of product toxicity through our electrically-driven cell-free monooxygenase system.

BREES - Yeasts' best friend

The easiest way to tackle the toxic effects of monoterpenoids on cells, is to circumvent the need for cells. The second component of MonChassis converts the already valuable monoterpenoids from our yeasts into even more valuable molecules. MonChassis utilizes the purified CYP BM3 from the bacteria Priestia megaterium (formerly Bacillus megaterium). However, using cell-free systems has its limitations. Traditional cell-free biocatalysis requires the addition of cofactors, such as NADPH that raises costs for large-scale applications (Bell et al., 2021). In order to make MonChassis in vitro component applicable at an industrial scale, we cut the need for expensive cofactors out of the equation by using electricity as a cheap alternative. Therefore, we created BREES, a BioReactor for Enzymatic ElectroSynthesis. We aimed to build a modified BM3 that can utilize electrons directly from electricity and does not require the addition of expensive cofactors (Fig. 5). We added a peptide-linker to the BM3 that can immobilize the enzyme to an indium tin oxide electrode and enable direct electron transfer from the electrode to the enzyme (Zernia et al., 2018). This drastically decreases the operating cost of MonChassis in vitro component compared to conventional cell-free biocatalysis. In combination with the engineered yeast strains of MonChassis, BREES allows the rapid, flexible, and large-scale production of monoterpenoids by overcoming all current limitations of traditional production methods.

Verbenone – Finding the needle in the haystack of endless possibilities

Seeing that monoterpenoids are a versatile class of substances and the possibilities of application are seemingly endless, we seek to find a suitable monoterpenoid to manufacture as proof-of-concept. For this, we focused our attention to local problems, where monoterpenoids could tackle the issue. A landscape development of special interest for us is the enormous deforestation of our local forests (e.g. the Sauerland wood) that has increased significantly, especially in the past year. (Interview with Jan-Otto Hake, forestry expert) A main contributor to this problem is the bark beetle, specifically the European spruce bark beetle that causes immense damage when appearing in large numbers. The beetles tunnel into the bark, cut off the water supply and consequently deaden the tree. Tying this together with MonChassis, we decided to produce the monoterpenoid verbenone, a pheromone that can be used as a bark beetle repellant.
The problem of bark beetle infestations and the resulting deforestation is not only a local problem but manifests in the entire northern hemisphere. Approximately 14.5 million cubic meters of timber were affected by bark beetle outbreaks in Europe between 2002 and 2010 (Sommerfeld et al., 2021). The climate crisis contributes to this threat, leaving trees even more vulnerable due to drought and its effect on the tree’s defense system, windthrow, and wood fires. Consequently, bark beetle outbreaks are predicted to worsen in the coming years, endangering entire ecosystems.
To combat outbreaks, several methods have been used. In some countries, large-scale commercial harvesting of infested trees or salvage logging has been done (Müller and Job, 2009). Another control method is the usage of pesticides to protect the lying wood that was not transported out of the forest in time from becoming a breeding ground. Others have decided to let the outbreak unfold with little intervention, e.g., the Bavarian National Park Forest.

As mentioned above, monoterpenoids are a substance class with many applications. An example of current relevance is the usage of verbenone against the spruce bark beetle (Fig. 6). Verbenone acts as an anti-aggregant pheromone, signaling other bark beetles that the tree is already occupied. During interviews with Dr. Dr. Lobinger, an expert on verbenone and bark beetles, she pointed out that the current production of verbenone is inefficient since the precursor of verbenone, α-pinene, is currently extracted from plants. This appears to be no reliable and sustainable solution to produce verbenone. Furthermore, we learned that commercially available verbenone suffers from significant batch effects that makes research on the deployment of verbenone inefficient.

Thus, verbenone is an ideal target to prove MonChassis’s capability to produce a monoterpenoid, which hangs far behind its potential because of unreliable large-scale production methods. For the BM3-catalyzed oxidation of α-pinene to verbenone, however, the BM3 shows low yields for the production of verbenone (S C Lehmann, 2016). Hence, we aimed to identify an optimized BM3 variant.
We decided to develop a protein engineering strategy to increase the verbenone formation rate of BM3. For an efficient screen of a large sequence space, we focused on establishing a high-throughput assay for detecting BM3-catalyzed verbenone formation. Two aspects are essential for a high-throughput assay: An easy readout of the product formation rate and no laborious purification of the protein variants.
The first issue is tackled by a colorimetric detection method for verbenone, which we developed based on verbenones reaction with the cheap reagent 2,4-Dinitrophenylhydrazine (DNPH). Upon reaction, it forms an orange hydrazone that can easily be quantified by absorbance. Concerning the purification of the BM3, we relied on the well-established autodisplay system, where single BM3 variants are presented on the outer membrane of Escherichia coli cells (Jose and Meyer, 2007). Combined with the colorimetric detection method, this is the first assay that allows quantifying verbenone in a high-throughput manner. Therefore, we build the optimal foundation for subsequent engineering campaigns of the BM3 for higher verbenone production rates.

Human practice

The contribution to a more sustainable bioeconomy is equally important to us, as preserving the forest and the adequate consideration of socio-ecological consequences concerning technological transformations. In order to have a deeper insight into these topics, we interviewed several experts in different fields. These, among others, include an expert on the usage of verbenone, a specialist on environmental law, and a potential end user.
More information regarding our human practices can be found on our human practice page.

The RFC1000-compatible yeast shuttle vector set

The decision to use the RFC1000 assembly standard was based on our approach to test many different metabolic designs for our metabolic pathways in yeasts. The pSB1K0X and pSB3C0X vector series of the iGEM registry, however, are not adapted to the expression of transcription units in yeasts. Thus, in cooperation with the iGEM Team TU Dresden, we developed a yeast shuttle vector set that will significantly simplify the work with S. cerevisiae for all upcoming iGEM teams.
More information regarding our shuttle vectors can be found on our partnership page.

Thymol synthesis

The iGEM Team TU Dresden focuses on the development of a hydrogel system applicable for the treatment of chronic wounds. Since many monoterpenoids are already used as pharmaceuticals, we figured that the monoterpenoid thymol could be embedded into their hydrogel. Thymol has anti-inflammatory properties and is therefore promising in the treatment of chronic wounds.
More information regarding our thymol synthesis and our partnership with the iGEM Team TU Dresden can be found on our partnership page.

Modeling

An integral part of MonChassis focuses on the scale-up of monoterpenoid production. The scale-up of fermentation processes is critical for commercialization and is mostly constrained by operation costs for growth media or oxygen transfer. For a reduction of these costs and to increase the performances of our production strains we optimized the growth media by in silico simulation with 125 different carbon sources. A further approach for MonChassis was the identification of potential genetic engineering targets by applying the flux scanning based on enforced objective flux algorithm. Thus, we are able to build the perfect foundation for the scale-up and potential continuation of MonChassis with the help of computational modeling.
More information regarding our modeling can be found on our modeling page.