Design and in vivo assembly of switchable protein-protein interactions for transcription regulation

Many aspects of biology rely on molecules coming together in response to specific signals. An important example of this is gene regulation. The vast majority of cells in our body contain the same DNA, but they turn on specific genes in response to the signals that they receive. When genes are switched on or off inappropriately the behaviour of cells can change, potentially giving rise to diseases such as cancer or diabetes. Systems that enable us to turn genes on and off artificially would be powerful tools that could be used in a wide range of medical and biotechnological applications.

Genes are regulated by proteins called transcription factors (TFs), which bind to specific sites on DNA, and either turn genes on by recruiting other proteins, or turn genes off by blocking the binding of such proteins. Many of these TFs are switches that respond to small molecules. For example, a bacterial TF called Lac repressor controls a gene involved in breaking down sugar. When the sugar is present a small molecule binds to Lac repressor, changes the shape of the protein and turns the gene on. Proteins like the Lac repressor are used as tools to control gene expression in engineered cells, such as bacteria that have been altered to produce human proteins. However, our ability and desire to engineer cells with increasingly complicated biochemistry means that there is now a pressing need for new switchable TFs. Ideally, these would not interfere with normal cell biology, referred to as being orthogonal; they should have predictable and tuneable properties, leading to a reliable set of biological parts that can be adapted for different uses; and they should be controlled by small non-toxic molecules that can enter cells, and could therefore potentially be used as drugs.

We have developed rules that allow proteins called coiled coils to be designed and synthesised in the lab. Coiled coils can be designed to assemble in different ways; e.g., to bring together 2, 3, 4 or more proteins, which can bind to each other tightly or weakly. We have shown that these designed coiled coils can be used inside cells to bring together the proteins needed to control a gene. By altering the strength of the interaction between the components of the coiled coil we can control how much the targeted gene is turned on or off. These new TFs have predictable and tuneable properties, but they cannot currently be controlled by small molecules to act as a switch. In the proposed work we aim to build on our findings to produce coiled coils that assemble only in the presence of a small molecule.

To do this, we will "hollow out" the centre of a coiled coil (i) to weaken the interactions that normally hold it together, and (ii) to generate a space for small-molecule drugs to bind. Under the right conditions, the small molecule will tightly and specifically fill the gap created, and will act as the missing piece of the jigsaw to allow the coiled coil to form and bring together the proteins needed for gene regulation. To find such conditions, we will test a large and diverse range of coiled coils that have hollowed cores of different shapes, sizes and chemistries. We will also test a diverse sample of potentially complementary small molecules. For this reason, the work is being done in collaboration with AstraZeneca, who have large libraries of small molecules, and the expertise to guide us towards the most promising of these.

The outcome of this project will be a series of compact, well-understood coiled coils whose assembly inside or outside of cells can be controlled by adding a small molecule or drug. The work will also develop the knowledge and procedures needed to make new switches in future. We will use these systems to produce TFs that can turn genes on and off in both bacteria and in human cells, and with Medimmune we will apply these tools to current problems in the production of biological molecules of medical interest.

This project will develop ligand-inducible protein-protein interaction domains based on rationally designed coiled coils, and apply them to the control of gene expression in bacteria and human cells.

The ability to control protein interactions is a powerful tool for the manipulation of cells and the design of biological systems. Switches based on natural inducible proteins have been used widely, but to implement the complex biological designs arising from current research we must find new ways for protein-based systems to sense and respond to environmental signals. Ideally, these systems would have well-defined, adaptable and reproducible characteristics, and would not interfere with the cellular machinery. Developing systems based on synthetic peptide sequences that are designed from first principles offers the scope to tune affinities and other properties to particular applications, and to introduce sensitivity to small ligands that are orthogonal to normal cell biochemistry.

In the proposed project, we will generate 4-helix coiled coils whose assembly and stability depend on the binding of a small ligand at the core of the complex. We will first design and characterise novel 4-helix bundles that are able to bring together components of transcription complexes within bacterial cells. We will then destabilise these bundles by substituting side chains in the core, producing libraries of partly hollowed-out variants with chemically diverse cavities. We will screen for protein/small molecule combinations that give coiled-coil assemblies only in the presence of complementary small molecules.

We will apply the new ligand-inducible coiled-coil domains to the control of gene regulation in bacteria and mammalian cells. The project will deliver regulatory components for the control of gene expression in basic research and biotech applications, and design principles to enable the further development of de novo designed ligand-responsive proteins for broader use.

INDUSTRIAL BIOTECHNOLOGY
Following on from above, industrial biotechnologists would benefit from more-programmable transcription factors and improved control of gene regulatory networks. To give our approach the best chance of success in forging new application areas in biotechnology, this proposal is a collaboration between the University of Bristol, Medimmune and AstraZeneca. This will ensure that tools and knowledge arising will be applied immediately to R&D projects, including the controlled production of biologics in wide mammalian-cell contexts. This collaborative translational approach would enable other application areas to be developed with our partners, by other academics, and by the biotech sector more generally. For example, one possibility that the proposed strategies and resulting materials would offer is the potential to develop biosensors that report directly on the presence of specific target molecules (drugs, toxins etc.) within complex mixtures; i.e., the protein-protein interaction libraries and screens could be made to respond to and, therefore, to sense small-molecule analytes.

More generally, we foresee the following three areas of impact from the proposed research.

NEW WAYS OF WORKING WITH WIDER ACADEMIC AND INDUSTRIAL COMMUNITIES and TRAINING
As noted above, a key challenge for protein design and synthetic biology is to translate fundamental research into real-life applications. To do this, we must train current researchers in academia and industry as well as early career researchers to work together creatively and productively. It is for these reasons that we have chosen from the outset to work with industrial partners (MedImmune and AstraZeneca) to conduct this pre-competitive research. This relationship will benefit our partners by exposing them to cutting-edge basic science in protein design and transcription control. It will benefit us by focusing our basic research, once achieved, on relevant challenges to the biotech sector. Moreover, it will benefit the 2 PDRAs on the project, allowing them to conduct basic research in a collaborative context where the real-life applications and industrial pull are clear. This will be invaluable experience for them in their future careers.

TRANSLATION
As noted above, there is a genuine ambition to see the proposed work through from basic science to real applications in biotechnology and, in particular for the production of proteins of medical importance and of benefit to MedImmune's core mission. Therefore, together with Letters of Support from MedImmune and AstraZenaca, the University has agreed a Heads of Terms with these partners. This covers industrial contributions, IP management, publication policy and so on. This demonstrates a clear intent to collaborate across all aspects of the work, i.e., from the basic underpinning science through to in-house projects at MedImmune that best facilitate translation.

OUTREACH
Following on from the latter, the proposed work seeks to produce tools that can be used in the short-to-medium term to improve the production of medically or biotechnologically relevant proteins. Thus, in the medium-to-long term, the public could benefit from this work through the availability of products or treatments that could not otherwise be obtained, or that could be produced more cheaply or efficiently through application of the tools and methods that we develop. As a result, and whilst the intricacies of the basic research might not be of interest to general audiences per se, the potential application areas for it almost certainly will be. The academic PI and Co-I have longstanding records in public engagement around bionanotechnology and synthetic biology. They will continue to raise public awareness and understanding of issues around and the potential offered by synthetic biology and biotechnology, and to encourage school pupils and young adults to pursue scientific education and consider scientific careers.