Proposal for GenScript grant

System based on CRISPR-Cas9

2016 EPFL iGEM team

Summary

CRISPR-Cas9 has already revolutionized synthetic biology. To build forward on this development we aim to implement digital-like circuits in yeast using a dCas9-based reprogrammable transcription factor. Recently, a study published the use of the modular CELLO software automating the design of DNA circuits using transcription factors in E. coli. As a proof of concept we will modify CELLO to use our dCas9 transistors in yeast for a so-called half-adder system, using AND and XOR gates, that we can then experimentally assess. With this approach we pave the way for even more complex biological circuits in yeasts.

Full description 1. Background

Coordination of vital functions in our body is dependent on intricate circuits of transcription factors (TFs) acting on gene regulation. In the early 2000’s, scientists took interest in reproducing those circuits in simple in vivo models, such as E. coli, to reprogram them and change functionality. Until recently, this was done by concatenating different TFs found naturally in the cells (Nissim et al., 2014).

Synthetic biological circuits allow gates - analogous to logic gates used in electronics - to be used in combination to create complex functions within the cell that can be receptive to external input. Each gate performs a simple calculation with given inputs and gives an output, which can be processed by other gates, acted upon by the cell, or observed by a researcher. A gate performs a logical function on one or more logical inputs. For example, with just one gate, a researcher can make a cell produce yellow fluorescent protein (YFP) if molecules A and B are both present within it. Another gate might give a positive signal in the presence of one or the other. With two gates, that cell can produce YFP if molecule B is present, but not A. With the emergence of the open-source software CELLO, recently published in Science, this process is significantly simplified and far less time consuming. CELLO is an extraordinarily powerful tool because it allows researchers to design circuits with up to sixteen gates in E. coli, allowing cells to investigate complex molecular pathways, or engineer cells that can be used in therapy for diseases. The process is entirely automated, taking only the desired circuit as input. CELLO assembles a plasmid containing this circuit using a library of sequences that code for TFs, promoters and other biological parts that together will be able to perform as a circuit. This plasmid, once constructed and transfected into a host organism, will recreate the desired circuit. Currently, CELLO only uses promoters that can be repressed using small metabolites to create its gates. Each one of these promoters and gates needs to be characterized, meaning that its output needs to be predictable given a certain level of input. Notably, gates for CELLO have only been characterized in E. coli thus far. This means that the most complex gates that can be created can use a maximum of sixteen

gates, but only in E. coli.

2. Research Description

We aim to expand on this theme, and hopefully make this powerful tool even more so. The first modification we would like to introduce is a dCas9-based system, already partially designed and constructed by the previous iGEM team from the EPFL. In this system, a catalytically dead Cas9 protein is bound to a VP64 activating subunit and essentially acts as a reprogrammable TF. This dCas9 can be integrated into the host genome fairly easily. This could provide numerous benefits to the program. In our system, the small guide (sg)RNA would target a 23-nucleotide sequence that would be predicted by the program and placed before the gate, forming an artificial promoter. Although these artificial “promoters” could not be activated or repressed by environmental signals, such as metabolites, they could form the basis of communication between intermediate gates in the circuit. This would free up the limited number of inducible and repressible promoters to be used for the initial gates, therefore greatly increasing the potential complexity of the circuits that could be created.

Using dCas9-based gates also has other benefits. sgRNAs are much smaller than the genes coding the TFs they might replace, which would reduce the space that each gate occupies on a plasmid. Very little toxicity has been previously observed with the dCas9-sgRNA complex, while TFs may create toxicity in cells. The combination of reduced space and reduced complexity further suggests that implementing our system could lead to more complex circuits being created automatically.

A second goal of our project is to create this system in S. cerevisiae, which means that all gates and promoters will have to be characterized in this organism. This would increase the power of CELLO, since it would allow its method to be applied in eukaryotic systems. Although this would represent a first step in this direction, we predict that this yeast-oriented program could be modified to be used in other eukaryotic systems with relative ease, since many of the genetic components that are used are the same or similar (the polymerase II, for example).

Modifying the program to include new gates in a new organism is fairly straightforward, and already underway. Luckily, CELLO is a modular program. This means that new data can be provided to it in a separate file, called a User Constraint File, and CELLO will use the same algorithms to interpret this new data and create plasmids that are suitable for the new organism and use the gates specified by the user. Altogether, constructing and characterizing these new parts will require a great deal of lab work, making this largely a wet-lab project. Once the parts are characterized, the data obtained can be put in the user-constraint file, completing the foreseen project.

In the case that we have extra time, we plan to review whether all the functions of the program are necessary, given the new Cas9-based system, and remove pieces that are no longer necessary. In addition, we would like to add an extra library to the project, which will add a further level of abstraction to CELLO. What this means is that researchers who are not familiar with how transcriptional circuits might work in a certain

organism might still be able to create circuits in that organism. Instead of telling the program to create a certain circuit, the researcher might simply say, “create a circuit that will produce YFP when the cell is in an environment that contains high levels of glucose and iron, but low levels of zinc.” The program would then look in our open- source library for known sensors for glucose, iron, and zinc and create the circuit itself.

Potential applications and implications of the project

Our project aims to increase the scope of genetic circuit automation to eukaryotes and enable this powerful technology to build more complex circuits. As a first proof-of-concept our team plans to construct a half-adder using the DNA sequence indicated by our modified CELLO. A half-adder permits the researcher to know when one molecule but not another one is present. The further applications of this project range from studying existing systems to creating novel ones. In yeasts, the complex interactions of metabolic processes could be studied by having different fluorescent proteins produced when certain proteins are abundant, and others produced when they are low. This could be used to give real-time data about the whole system at once. Once finished, this project could be easily redirected into other eukaryotes as well, and could have interesting therapeutic applications. For example, novel behaviors could be programmed into cells based on the environment they are in, which could be identified by the amount of secondary signals coming from receptors. In this example, researchers could reprogram cells to produce a therapeutic protein when in a certain niche, and this application would be straightforward using our project.