Team:Toulouse INSA-UPS/Engineering
Overview
Synthetic biology is often described as the design and construction of new biological elements, devices, and systems, as well as the redesign of existing natural biological systems for application purposes (Nature, 2021). As in any other engineering field, synthetic biologists follow iterations of the DBTL cycle (Design, Build, Test, Learn) to successfully develop innovative biological systems (Figure 1).
Figure 1: Iterative design-build-test-learn (DBTL) cycle in Synthetic Biology (Waldby et al. 2018).
In the framework of our project ELIXIO, we were able to design various genetic
constructs,
successfully assemble them, insert them into their final chassis and finally
demonstrate the
production of three of the main molecules of the violet scent.
While these results may seem very positive for a team of only six students and in a time
frame as short as four months, this engineering success was not always as straight-forward
as presented in our Results section.
To illustrate this, we have chosen to detail here the example of our pVIOLETTE
construction
for the production of α-ionone, which proved to be a significant challenge at
various
levels. In the end, four iterations of the DBTL cycle had to be performed before
reaching
the final proof of concept.
Cycle 1: Designing pVIOLETTE and first cloning attempts
Design
α-ionone is produced in two steps from lycopene, it is first converted to ε-carotene (lycopene ε-cyclase activity) which is transformed into α-ionone (carotenoid cleavage dioxygenase activity). A previous study showed that for the first step, LcyE from Latuca sativa and for the second step, the CCD1 from Osmanthus fragrans (OfCCD1) could be used to produce α-ionone (Chen et al. 2019). When researching the literature on the production of α-ionone, it was found that the second step was rate limiting (Chen et al. 2019; Werner et al. 2019). The main hypothesis to explain that is the poor accessibility of the OfCCD1 enzyme, which is cytoplasmic, to its substrate which usually is trapped inside the membrane. To address this, OfCCD1 was fused to the membrane-bound LcyE in a previous study in E. coli to facilitate substrate channeling. This strategy allowed a >2.5-fold enhancement of α-ionone in E. coli (Chen et al. 2019). Inspired by this work, we chose to design the same fusion enzyme and codon-optimize it for expression in S. cerevisiae to see if this strategy could also work in our yeast (Figure 2).
Figure 2: Final design of pVIOLETTE.
Build
The different gene fragments were ordered from IDT and the cloning vector was provided to us by Sara Castaño Cerezo. There were three different gene fragments to assemble in the vector:
-
Nourseothricin resistance gene expression cassette (NsrR hereafter)
-
PGAL1 promoter, LcyE gene and half of the LGS linker (LcyE hereafter)
-
Other half of the LGS linker, OfCCD1 gene and TVPS13 terminator (OfCCD1 hereafter)
The different fragments as well as the cloning vector were amplified by PCR. It was first tried to assemble the construction in one step through InFusion cloning. Even if the positive cloning control worked, no clone was ever obtained on this cloning. After trying several times and changing experimenters, it became clear that this strategy would probably not work.
Learn
Our hypothesis was that attempting to clone three inserts into a vector in one step was perhaps a bit bold.
Cycle 2: Two steps cloning strategy
Design
Because of these results, it was decided to attempt to clone pVIOLETTE in two steps. In the first step, the two inserts OfCCD1 and NsrR would be assembled in the vector (this intermediate plasmid was named pVIOLET). In a second step, the last LcyE insert would be inserted to finally have the final construct pVIOLETTE (Figure 3).
Figure 3: Two steps cloning strategy for pVIOLETTE.
Build
New primers were ordered, and the different fragments were again amplified by PCR. The
intermediate construct pVIOLET was successfully assembled and verified by performing a
restriction profile and by sequencing.
For the cloning next step, however, the various tests never resulted in positive clones.
Because of this observation, it appeared that an inherent feature of the LcyE block sequence
made cloning impossible.
Learn
We analyzed the sequence of the fragment and realized that the LGS linker had a very high GC content compared to the rest of the sequence. According to the Takara Infusion kit handbook, a GC-rich homology zone is less likely to work for InFusion cloning.
Cycle 3: Ordering a new gene fragment to move the homology zone out of the linker
Design
Based on this observation, it appeared that our whole cloning strategy had to be changed again. A new 500 bp fragment (NewLinker) flanking the linker was designed together with new LcyE (NewLcyE) and OfCCD1 (NewOfCCD1) fragments (Figure 4). The previous pVIOLET intermediate plasmid was used as template for an inverse PCR with the aim to obtain the linear vector with the NsrR cassette (new vector named pVIO).
Figure 4: New cloning strategy with the final homology regions out of the linker.
Build
Then an InFusion was made with the three insert NewLcyE, NewLinker and NewOfCCD1 with the
linear vector pVIO. With this strategy, the homology regions were not in the GC-rich linker
anymore. This times, many clones were obtained after incubation of the cloning plates.
After, screening the clones with colony PCR, performing restricting digestion profiles and
sequencing the positive clones, it appeared that this strategy was successful.
The next step was to integrate this vector in the genome of our parental yeast strain
LycoYeast. We tried three different yeast transformation protocols, tried changing
experimenters, but were never able to obtain yeast mutants.
Learn
After discussion with yeast experts, it appeared that to obtain chromosomal integration of our construct into the S. cerevisiae genome, a much larger amount of DNA had to be put in to hope to obtain mutants. Even though in the yeast transformation protocols we used, the amounts of DNA used for transformation rarely exceed 1 μg, they told us to use as much as possible to increase our chances.
Cycle 4: Construction of the LycoYeast-VIOLETTE strain
Build
With this new information in mind, we tried transforming LycoYeast by adding 5μg of linear pVIOLETTE in the transformation mix. This times, many clones were obtained and the correct integration was checked by PCR. This new cycle of engineering therefore allowed us to finally obtain our LycoYeast-VIOLETTE strain.
Test
The next step was to finally test our strain. To do see, we cultivated it in a galactose
containing medium (YPGal) which is supposed to induce the expression of our fusion enzyme.
We noticed that in the induced cultures, a color change was observed compared to the
parental LycoYeast. This was expected since lycopene is red while carotenes are orange. We
were also delighted to find that a characteristic ionone sweet smell could be noticed!
To confirm these observations, a rigorous methodology of measurement was applied with the
development and application of analytical methods of carotenes in HPLC and ionones in GC-MS
(Figure 5).
Figure 5: Strain characterization of LycoYeast-VIOLETTE. Upon galactose-induction, lycopene is converted to ε-carotene (HPLC data on the left) which is then converted to α-ionone (GC-MS data on the right).
Learn
These results allow us to conclude that the enzymatic fusion that we have designed BBa_K3930024 is functional in S. cerevisiae, which could serve as an avenue to improve the production of α-ionone by biotechnological means in this organism.
Conclusion
We approached this project with the same optimism and big ideas that all iGEMers have. Our instructors repeatedly warned us that most experiments are doomed to fail. Engineering a biological system is indeed a complex, challenging, and often slow process. Here we have described, using the example of α-ionone, how by performing successive iterations of the DBTL cycle we were able to create a violet-scented yeast (Figure 6).
Figure 6: Summary of the different engineering steps completed before reaching the final proof of concept for the production of α-ionone. D=Design, B=Build, T=Test, L=Learn.
References
Chen X, Shukal S, Zhang C. 2019. Integrating Enzyme and Metabolic
Engineering Tools for Enhanced α-Ionone Production. J Agric Food Chem. 67(49):13451–13459.
doi:10.1021/acs.jafc.9b00860.
Waldby C, Gray P, Griffiths P, Vickers C, Meek S, Small I. 2018. Synthetic Biology in
Australia: an outlook to 2030. Australian Council of Learned Academies Expert Working Group
Horizon Scanning Project.
Werner N, Ramirez-Sarmiento CA, Agosin E. 2019. Protein engineering of carotenoid cleavage
dioxygenases to optimize β-ionone biosynthesis in yeast cell factories. Food Chemistry.
299:125089. doi:10.1016/j.foodchem.2019.125089.
