EVO-ENGINE: A directed evolution engine for engineering proteins and logic gates

Lead Research Organisation: Imperial College London
Department Name: Life Sciences

Abstract

Natural selection is the most powerful force in biology and drives the evolution of complex organisms, such as ourselves, simply by continually enriching for those that survive and replicate. This has been long exploited in the field of "directed evolution" which copies natural selection in an artificial manner, to obtain useful proteins. Directed evolution allows us to make proteins with complex properties that would be impossible to achieve by design alone. Perhaps the best-known example is the now widely-used cancer drug, Herceptin, which was originally evolved in the laboratory using a technique called "phage display". This method uses tiny viruses called phage - which only infect bacteria - to evolve new proteins. Randomised proteins are made by the phage and only ones which stick to a desired target are retained and survive - and so they are said to be evolving. At the end of each experiment, the proteins are recovered and can be made in large quantities and purified. Around 20 drugs made by artificial selection (phage display) are already approved for use in humans or are in in late stage clinical trials. This technology is therefore very powerful and has already benefited humanity greatly.

It is possible to use phage and directed evolution not just for making proteins with new "sticky" properties but for engineering complex functions. Moreover, there is a new scientific field called synthetic biology which aims to re-engineer biological networks and systems to make useful enzymes, molecules and even small machines - such as biological computers and selective cancer cell modulators. One branch of synthetic biology aims to generate genetic circuits - similar to the logic gates that are used in standard electronics and computers - and these are currently made very laboriously, one-by-one, with a huge amount of trial-and-error. By applying directed evolution we believe that we can save a lot of time and effort by simply evolving logic gates in phage and bacteria, in order to obtain a new generation of useful biological tools. We have already built a prototype robot (The EVO-ENGINE) that grows phage and bacteria under a series of selective pressures. In this way, only the ones containing a functional logic gate survive. Over time, with random mutations, and increasing selection pressures, better and better logic gates evolve. The resulting components have a vast range of biotechnology applications, such as making biological sensors (for example, to detect pollutants in the environment) or for controlling synthesis circuits in bacteria to make useful products (plastics, fuels and medicines). We are only beginning to harness the power of directed evolution - which nature has exploited for millions of years - and this BBSRC-funded project has the potential to provide a major advance.

Technical Summary

Directed evolution is a powerful technology that has been used in protein engineering to make new proteins, such a widely-used peptide cancer drugs. By applying a Darwinian selection pressure to mutating proteins or gene networks, so that only the desired variants survive, much larger combinatorial spaces can be processed compared to screening rationally-designed constructs one-by-one. Despite this potential, directed evolution has been relatively underused in the expanding field of gene circuit engineering, and most projects have focused on engineering logic gates, or other functions, by iterative design-testing cycles.

To exploit this potential, we have made a custom-built programmable evolution robot (EVO-ENGINE) that harnesses bacteriophages and bacteria to allow the evolution of arbitrarily complex multi-input genetic logic functions. In supporting preliminary work, we have generated the first set of orthogonal dual activator-repressor TFs to work on complex multi-input promoters in bacteria. In this project, we will go beyond this and provide the first demonstration that logic gates can be evolved and optimised on a directed evolution platform. We will exploit the new TFs that we have already engineered as a starting point (based on bacteriophage lambda cI), and will build NOR and NAND logic gates. These are particularly useful because they are 'functionally' complete. In electronics, by layering these gates, it is possible to obtain all other logic functions. Our system is based on modular, orthogonal (non cross-reacting) transcription factors and so naturally lends itself to downstream expansion and layering, for increasingly complex biological computing applications.

Our system has several advantages over standard directed evolution, primarily in terms of handling vast combinatorial sequence space. Thus, we will ultimately use the system in to explore the evolution of new proteins de novo, which is currently at the frontiers of directed evolution.

Planned Impact

The new discipline of synthetic biology - where engineering principles are applied to biological systems - promises to add a new layer of genetic control to metabolic engineering that will greatly impact both biotechnology and society.

Gene network engineering, based on the EVO-ENGINE, and its capacity to evolve logic gates and other genetic functions or proteins, will impact the biotechnology industry, where genetic and metabolic networks are already being used to make products of industrial and commercial value.

For example, gene networks to improve the production of biologics (pharmaceuticals), high-value chemicals, and genetically modified organisms are already being implemented in bacteria. These will be impacted in the future by the development of functional, modular, orthogonal, expandable, functionally-complete logic gates that we are carrying out in this project. Thus, in the longer term, is likely that biotechnology companies will be able to translate this basic biological engineering knowledge into new products, based on intelligent synthetic networks, regulated by logic.

The engineering of proteins will have a significant impact in biological manufacturing and the optimisation of development of novel high-value chemicals by fermentation methods, which will have an impact in industrial biotechnology and bioenergy.

Though our regular outreach efforts (e.g. Imperial Festival open day, science cafes, open lectures, etc.) we will increase the public awareness of bioreactors, which will contribute to the engagement of citizens in science.

Lastly, it is important to consider the impact that this project will have in terms of training of young scientists in a new interdisciplinary field - synthetic biology. We will engage students to participate at the iGEM competition to exploit the societal and economic impact of this research. Furthermore, as well as training the specific project postdoctoral staff in a hybrid of computation, robotics, logic and wet biological engineering, the resulting scientific publications are likely to be very influential in training a new generation of scientists in the nascent field of gene circuit engineering.
 
Description We have built a directed evolution robot (Evoengine) and have engineered the smallest activator transcription factor ever described ("Accelerated evolution of a minimal 63-amino acid dual transcription factor" Science Advances 6: eaba2728 (2020)). These tools will have important implications in bioengineering of artificial gene circuits.
Exploitation Route We have deposited phagemid directed evolution tools for use in the Evoengine framework. These are freely available at: https://www.addgene.org/Mark_Isalan/ . To date, Addgene has distributed 105 sample(s) on our behalf to the research community (Feb 2021). Thus our work is being put to use by others.
Sectors Healthcare

Manufacturing

including Industrial Biotechology

Pharmaceuticals and Medical Biotechnology

URL https://advances.sciencemag.org/content/6/24/eaba2728.full
 
Title Deposited plasmids on Addgene 
Description We have deposited phagemid directed evolution tools for use in the Evoengine framework for access at: https://www.addgene.org/Mark_Isalan/ These tools represent the work from three main papers funded here: 1. Brödel AK, Rodrigues R, Jaramillo A, Isalan M, Engineering the smallest transcription factor: accelerated evolution of a 63-amino acid peptide dual activator-repressor. Science Advances 6: eaba2728 (2020). 2. Brödel, A.K., Jaramillo, A. & Isalan M. Intracellular directed evolution of proteins from combinatorial libraries based on conditional phage replication. Nature Protocols Sep;12(9):1830-1843. (2017). 3. Brödel, A.K., Jaramillo, A. & Isalan M. Engineering orthogonal dual transcription factors for multi-input synthetic promoters. Nature Communications 7, 13858 (2016). 
Type Of Material Biological samples 
Year Produced 2017 
Provided To Others? Yes  
Impact We have deposited phagemid directed evolution tools for use in the Evoengine framework. These are freely available at: https://www.addgene.org/Mark_Isalan/ . To date, Addgene has distributed 105 sample(s) on our behalf to the research community (Feb 2021). 
URL https://www.addgene.org/Mark_Isalan/
 
Title Directed evolution robot 
Description We have built a directed evolution robot and have engineered the smallest dual transcription factor ever described (Science Advances 6: eaba2728 (2020)). These tools will have important implications in bioengineering of artificial gene circuits. We have published a lead-up paper for the wet biology technologies used in the robot: Brödel, A.K., Jaramillo, A. & Isalan M. Intracellular directed evolution of proteins from combinatorial libraries based on conditional phage replication. Nature Protocols Sep;12(9):1830-1843. (2017). 
Type Of Material Technology assay or reagent 
Year Produced 2020 
Provided To Others? Yes  
Impact Using the new technology,we have engineered the smallest activator transcription factor ever described (Science Advances 6: eaba2728 (2020)). These tools will have important implications in bioengineering of artificial gene circuits. 
URL http://doi.org/10.1101/725739
 
Description Imperial-Warwick-CSIC Valencia Collaboration 
Organisation University of Valencia
Country Spain 
Sector Academic/University 
PI Contribution This grant has allowed the teams of Prof. A Jaramillo (Warwick Uni/CSIC Valencia) and Prof. M. Isalan (Imperial) to establish a fruitful collaboration at the interface between physics, robotics and synthetic biology for machine learning/AI inside bacteria.
Collaborator Contribution We have built a directed evolution robot and have engineered the smallest activator transcription factor ever described (Science Advances 6: eaba2728, 2020). This work has also allowed us to apply for follow on funding for the new field of biomachine learning. We now have a breakthrough paper: "Engineered gene circuits capable of reinforcement learning allow bacteria to master gameplaying." Bacteria capable of AI/ML teach each other to master games such as tic-tac-toe: https://doi.org/10.1101/2022.04.22.489191
Impact Brödel, A.K., Jaramillo, A. & Isalan M. Intracellular directed evolution of proteins from combinatorial libraries based on conditional phage replication. Nature Protocols Sep;12(9):1830-1843. (2017). Brödel, A.K., Isalan M. & Jaramillo, A. & Engineering of biomolecules by bacteriophage directed evolution. Current Opinion in Biotechnology 51, 32-38 (2018). Brödel, A.K., & Isalan M. Trp-ing upon synthetic repressors. Nature Chem Biol 14 328-330 (2018). Brödel AK, Rodrigues R, Jaramillo A, Isalan M, Engineering the smallest transcription factor: accelerated evolution of a 63-amino acid peptide dual activator-repressor. Science Advances 6: eaba2728 (2020). Engineered gene circuits capable of reinforcement learning allow bacteria to master gameplaying. https://www.biorxiv.org/content/10.1101/2022.04.22.489191v2.full (2023)
Start Year 2017