Team:GO Paris-Saclay/Implementation
PROPOSED IMPLEMENTATION
Our project HuGenesS is based on the entanglement of 2 genes within the same nucleotide sequence so that they totally overlap but are read in two different reading frames. Thanks to the software CAMEOS published only a year ago, it is now possible for synthetic biologists to design entangled genes for many types of applications. Therefore, HuGenesS represent a new synthetic biology tool with tremendous potential for all organisations working with biological products with high added value, for example biofuels, bioplastics, vitamins and antibiotics production (Najafpour et al. 2007), biocatalysts (Paul et al. 2019)… This field is increasing so rapidly that it is estimated that bio-based chemical sales would grow about 15% in the next five years (Paul et al. 2019).
Anchoring HuGenesS in the real world through our human practices
In order to anchor this new technology in the real world, the first step was to discover and understand in depth the mechanisms of CAMEOS. Personal research was carried out in the CAMEOS system, then we were able to compare our point of view with the authors of the article introducing this tool, but also with researchers specialising in bioinformatics, genetics or evolution (see page HP). The result of this thorough inspection is a detailed tutorial for a clear and simplified use of the method and its stakes
But the anchoring of a new idea in our society requires communication. Events such as meetings, the "iGEM Fr" popularisation video, the holding of a stand at the Cité des Sciences et de l'Industrie, enabled us to reach a large audience. Thus, we were able to confront all types of people, with their knowledge, their ideas, their fears, and thus explain to them the solutions we wanted to provide thanks to HuGenesS.
Finally, in order to make these first approaches concrete, we were lucky enough to be able to design a sequence for Dr. Tristan Rossignol, who as a researcher commissioned us to create a tangle for his personal work. This achievement is the first step towards the concrete applications that will become possible in the near future.
Identifying applications and opportunities for implementation
We envisage that our project could be useful to research laboratories, pharmaceutical laboratories, as well as cosmetics or biofuel companies. For indeed, their needs are diverse, but so are the solutions proposed by HuGenesS. This can be explained by the very essence of the programme we use: the CAMEOS (Blazejewski et al. 2019) software takes 2 distinct sequences as input to entangle them. It is therefore when choosing the sequences to be entangled that the potential of our project lies. We believe HuGenesS could be implemented for three main goals: to stabilize genetic information, to secure bioconfinement of GEMs and prevent horizontal gene transfer, and to increase genetic information in small genomes.
Stabilisation of genetic information
Genetically engineered organisms often experience a reduction in fitness that leads to genetic instability of the engineered functions. In most cases, this organism will naturally tend to lose this gene. A striking example is the case of cyanobacteria, which seems to have defence systems that enable it to inactivate exogenous genes (Rosenberg et al. 2008). This is a known phenomenon, and we know that if we imagine an infinite lifespan for these strains in incubators, at a certain moment the proportion of cells still carrying the introduced gene will be zero. In the short term, this requires a lot of maintenance, manipulation and selection of the strains so that the introduced gene is maintained in their genome. By implementing HuGenesS, we will couple a gene of interest to an essential gene of the organism. Therefore if a mutation occurs in this construction, it will be lethal to the cell. All living organisms will therefore be carriers of your gene of interest intact.
Implementation in pharmaceutical companies
Many pharmaceutical laboratories produce chemical molecules using bacterial or fungal cultures. The example of antibiotics such as the glycopeptide antibiotic (GPA) A40926 which is the precursor of dalbavancin (Yushchuk et al. 2020), or anti-cancer drugs (Madduri et al. 1998), simply illustrates the need for modified micro-organisms (Parekh et al. 2000) to express these molecules in a massive way. Indeed, natural micro-organisms are either unable to produce these molecules, or for the most part, not in these quantities. Moreover, these strains must survive this overproduction which can be toxic for them. On an industrial scale, these drug productions require litres of fermentation cultures. So the slightest loss of efficacy, due to a mutation for example, can be extremely expensive. Moreover, some natural gene clusters are lost in fermenter because their expression cost a lot of energy to the cell and does not remain essential to the cell to survive.
HuGenesS may be the solution: if we find a gene well adapted to the one artificially inserted to produce the molecule, then each micro-organism will necessarily keep its production function. In this way, pharmaceutical laboratories will not have any loss due to the randomness of mutations over time.
In short: a notorious saving in time and money thanks to the stabilisation of the introduced or natural genes in fermenters.
Implementation in biofuel companies
On current trends, biofuels are developing exponentially (overall increase of 55% in 2019 in France) and research in this field is at its peak. One of the major problems is that the production of these fuels from plants and algae is very expensive, and in the case of 1st generation biofuels, it requires a very large amount of agricultural land and water.
But let's talk about 2nd and 3rd generation biofuels. Here, micro-organisms become essential to their production. Indeed for the degradation of lignocellulosic waste (2nd generation), modified bacteria and fungi are the only organisms capable of this transformation. Similarly, the production of 3rd generation fuels using microalgae requires genetic manipulation in order to become viable. While the bio-containment of algae in the bioreactor is essential (Rosenberg et al. 2008), it also appears that the cyanobacteria used have a strong tendency to mutate the genes introduced thanks to systems for the inactivation of exogenes (Rosenberg et al. 2008). HuGenesS can become the ideal solution: Encapsulate the transgenic gene with an essential gene from the cyanobacteria, the production function will be preserved!
Implementation in the cosmetic industry
The field of cosmetics is in continuous development, always looking for better, more durable solutions to make humans beautiful and healthier. Trans-resveratrol is a plant phenolic compound with proven benefits for human health. Currently, the growing demand for trans-resveratrol cosmetics makes it necessary to produce it from sustainable sources. The adaptation of yeasts or bacteria with genes that code for enzymes of the trans-resveratrol pathway and the activation of plant cell metabolism would be one way to provide this production ( Donnez et al. 2009). Here again, these are genetically modified strains carrying at least some, if not all, of the sequences coding for the dedicated metabolic pathway. A heavy modification which would tend to be eliminated. HuGenesS solves these two problems: a new metabolic pathway, stable over time, and which imposes only a slight additional burden on the micro-organism.
Securing biocontainment of GMOs and preventing horizontal gene transfer
Securing self-killing mechanisms
Synthetic biologists have suggested gene circuits to increase bioconfinement. These include so-called kill switches that would be activated in conditions associated with an accidental release in the environment. One of these is the the Passcode Switch. This switch requires the combination of 3 different signals to prevent the production of a toxin. If the organism escapes from this very controlled environment, the toxin will be produced, and the GMO will die. But this genetic circuit contains an Achilles’ heel, ie the possible evolution of the toxin gene. Indeed, if mutations accumulate in the toxin gene or the organism loses it, then the confinement can be bypassed.
But if the toxin gene is entangled with an essential one, a mutation inactivating the toxin could have dire consequences for the gene essential for survival. It is what we want to propose with the entanglement of the CcdB toxin gene. Our solution is then a way to combine stability and security, to offer a better confinement.
Preventing horizontal gene transfer
Organisms can exchange genetic information from one strain to another, but also potentially from one species to another. This becomes a major problem when we design GMOs. In case GMOs are accidently released into the environment, we have to prevent their genetic modifications to become transferred to another type of organism. With HuGenesS, we can intertwine our gene of interest with a toxin gene in such a way that they become dependent on each other. For your designed GEM to survive, it will need to produce an antitoxin that would only be present/produced if the GEM is in its proper bioconfined environment (a fermenter for example).
The compaction of genetic information
Whether you want to secure or stabilise your construction, in all cases the added genetic information will be compacted. The genetic burden will be less and we can then imagine adding huge genes into very small genomes. Everyone will be able to see big inside smaller genomes!
The future in GEMs might rely in organisms with very small genomes, and bottom up approaches to synthetic microorganisms. Both are currently an active area of research (Liu et al. 2020). Although we did not explore it further, we believe HuGenesS could be a crucial help in these tasks.
Comparison to alternative biocontainement strategies to assess the potential market
Open the PDF file of our Analysis.
Improvement opportunities for HuGenesS implementation
In order to envisage such a prosperous future for our project, many improvements still need to be considered.
Streamlining pre- and post-CAMEOS analysis
The first critical step is to optimise the selection and sequence verification treatments to be carried out before and after their treatment by CAMEOS. Because if CAMEOS intertwines the genes, it is up to us to check that the sequences obtained are potentially usable and functional. In the tutorial course, we have made a course study to compare the multiple sequence alignments that can be generated as inputs. Improvements to select the variants generated by CAMEOS would be useful. Incorporation of 3D structure information would be advantageous, since certain amino acids are essential for protein structure and therefore for its function. The software has no way of taking these elements into account so far. Thus, as explained in the Model, the first optimisation step of CAMEOS could be modified so that it prohibits mutations in certain portions, thus ensuring that certain nucleotides and/or amino acids are preserved.
Furthermore, it has been observed that in nature, when the reading phase is changed from a coding phase, the theoretical peptide from this phase retains properties and patterns similar to those of the protein from the coding phase (Bartonek et al, 2020). An interesting study would be to test whether proteins with similar hydrophobicity profiles would yield better (ie functional) entanglements. This new approach could make it possible to understand why some entanglements appear better than others despite high identity scores. For this, a study of hydrophobicity profiles could be an important first step.
Adapting to the GC content and tRNA repertoire of the chassis
But we also want this new method to be adjustable to a wide variety of organisms, so it is important to be able to adapt to each of them. We know that the basic A/T and C/G proportions vary greatly from one species to another. Also we must make sure that the tRNA repertoire of our chassis will not be limiting for the codons that have been introduced in the HuGenes. These data should be considered by the program CAMEOS, in order to respect the existing proportions as much as possible.
Implementation possibilities in organisms with introns
Until now the software has only been used with genes from simple micro-organisms. If we want to consider use in plants and animals, we need to look at some fundamental differences, such as splicing. This process of removing introns from neosynthetic RNA allows the organism to have a gene encoding several proteins. But in our situation, this splicing of introns poses a real problem for us. It then becomes essential to know the exact splicing sites, as well as the signals recruiting the transcription and translation machinery for each gene of interest. We return here to the importance of optimising CAMEOS, as proposed in the modelling. And so, we could then imagine interlacing the second gene according to the same splicing (intron/exon) as the first, without having the certainty that each of these 2 genes would be expressed in the desired form. This improvement would be a considerable step forward for the future of CAMEOS.
Testing functionality of the entangled genes remains the main bottleneck
In the short six weeks we had in the laboratory during the summer, we were only able to clone (and characterize) one HuGenesS construct out of the twelve that were attempted. This construct, knt-gfp (BBa K3427001), encoded two proteins that appeared to have lost their function, a result that could be expected given the low identity scores of the modified proteins. However, even with higher scores, the functionality of the entangled genes needs to be tested experimentally, a process that constitutes the biggest bottleneck in full implementation.
Finally, the most important challenge to date is to succeed in informing potential users of such a bioinformatics possibility of containment and genetic stabilisation. It is only by communicating this novelty that HuGenesS will be able to propose advanced research and concrete improvements to make use of the immense potential of this technique.
The potential is there in the future to entangle more than two genes but to use all six frames!!!!
References
Parekh et al. 2000: Parekh, Vinci & Strobe (2000), Improvement of microbial strains and fermentation processes. Applied Microbiology and Biotechnology, doi:10.1007/s002530000403
Madduri et al. 1998: Madduri, Kennedy, Rivola, Inventi-Solari, Filippini, Zanuso, … Hutchinson (1998) Production of the antitumor drug epirubicin (4′-epidoxorubicin) and its precursor by a genetically engineered strain of Streptomyces peucetius. Nature Biotechnology, doi:10.1038/nbt0198-69
Yushchuk et al. 2020: Yushchuk, Andreo-Vidal, Marcone, Bibb, Marinelli & Binda (2020) New Molecular Tools for Regulation and Improvement of A40926 Glycopeptide Antibiotic Production in Nonomuraea gerenzanensis ATCC 39727. Frontiers in Microbiology, doi:10.3389/fmicb.2020.00008
Rosenberg et al. 2008: Rosenberg, Oyler, Wilkinson, Betenbaugh (2008) A green light for engineered algae: redirecting metabolism to fuel a biotechnology revolution. Current Opinion in Biotechnology, https://doi.org/10.1016/j.copbio.2008.07.008]
Donnez et al. 2009: Donnez, Jeandet, Clément & Courot (2009) Bioproduction of resveratrol and stilbene derivatives by plant cells and microorganisms. Trends in Biotechnology
Liu et al. 2020: Liu, Su, Li, Ledesma-Amaro, Xu, Du & Liu (2020) Towards next-generation model microorganism chassis for biomanufacturing. Appl Microbiol Biotechnol doi: 10.1007/s00253-020-10902-7.
Blazejewski et al. 2019: Blazejewski, Ho & Wang (2019) Synthetic sequence entanglement augments stability and containment of genetic information. Science, DOI: 10.1126/science.aav5477
Paul et al. 2019: Paul, Sangeetha, & Deepika (2019) Emerging Trends in the Industrial Production of Chemical Products by Microorganisms. Recent Developments in Applied Microbiology and Biochemistry, doi:10.1016/b978-0-12-816328-3.00009-x
Najafpour et al. 2007: Najafpour (2007) Production of Antibiotics. Biochemical Engineering and Biotechnology, doi:10.1016/b978-044452845-2/50011-2
Bartonek et al, 2020: Lukas Bartonek, Daniel Braun, and Bojan Zagrovic, Frameshifting preserves key physicochemical properties of proteins, PNAS March 17, 2020 117 (11) 5907-5912; first published March 3, 2020; https://doi.org/10.1073/pnas.1911203117