When we think of synthetic biology, we often think of engineering a cell to give it some useful function. But SEED 2016 had quite a few speakers working outside of a biological cell. Some broke open cells to utilize just the cellular machinery to create “cell-free” systems. Others showed what could be done inside of the computer (in silico) to improve our understanding and prediction of synthetic gene networks. Here, we’re highlighting SEED speakers who showed how both of these approaches can advance synthetic biology.
Cell-free synthetic biology
Roy Bar-Ziv gave the first keynote at SEED 2016. His group at the Weissman Institute has made tremendous progress toward using cell-free expression that can mimic the behavior of real cells. Over the last 12 years they developed their ‘artificial cells’ using microfluidics and DNA arrayed on 2D substrates as DNA brushes. Each spot of DNA can be programmed the same as DNA in cells, and unlike other cell-free expression setups the microfluidics allows for dynamics.
This work really matured in their 2014 paper in which they could even achieve oscillating protein expression in the artificial cells. 100s or 1000s of ‘cells’ with different DNA programs can be tested with dynamics determined by the diffusion of energy resources and protein products through different sized fluidic channels that connect cells and the energy buffer. He said that his goal is to create programmable cell-free systems from the gene, to the ‘cell’ to ‘multi-cellular’ communication.
— PLOS Synbio (@PLOSSynbio) July 18, 2016
Vincent Noireux, from the University of Minnsota, has been instrumental in pushing cell-free synthetic biology forward. He developed the TX-TL system used by many collaborators like Roy Bar-Ziv. On day 3 of SEED he presented his next generation cell-free TX-TL for synthetic biology.
He sees several advantages to cell-free systems: no interference or response from an organism, fast prototyping, quantitative concentrations that lend themselves to accurate mathematical models, and it’s safe since there are no living components.
— PLOS Synbio (@PLOSSynbio) July 20, 2016
The newest TX-TL can give 6-8 hours of expression ~2mg/mL of synthesized protein from a single mix that commercially available (MYtxtl from MYcroarray). CRISPR-Cas9 also functions in TX-TL, and they could even synthesize full T7, MS2, and ΦX174 phage). DNA replication also occurs in addition to the TX and TL of transcription and translation so they have the ambitious goal of creating synthetic cells.
In silico synthetic biology
Synthetic biology has always had a modeling component going back to the 2000 papers describing the genetic toggle switch or the repressilator. However, the field has really progressed to handle more of the nuances and details of the biological mechanisms of synthetic gene circuits.
Howard Salis presented the sequence-to-function algorithms that his group develops as design tools for the community. Coming largely from a biophysics and physical chemistry approach, he has developed several biophysical models that help design parts for synthetic biology. The most popular is RBS calculator, but there are also newer tools for more complex functions.One example is the operon calculator that works to maximize several of rules for a good operon design.
A unique question the Salis posed is whether RNA folding speed is faster than ribosome-RBS binding?
An informal poll of the room showed that most of us thought that mRNA folding would be much faster, but it turns out there are cases in which we are wrong. Throuhg a process called ribosome drafting, successive ribosomes bind faster than the mRNA’s refolding rate as they draft off of the ribosomes preceding them.
Can ribosome binding be faster than mRNA folding? Apparently yes when they "draft" off of each other. #SEED2016
— PLOS Synbio (@PLOSSynbio) July 18, 2016
Domitilla Del Vecchio talked about her work to mathematically model the context effects in living cells. Context refers to the changes in circuit behavior based on other circuit parts or changes in the cell. She had previously worked on the problem of retroactivity that changes the dynamics of gene circuit components as they connected. Now, she is working on the problem of resource sharing. Cells have a limited supply of resources like RNA polymerases or ribosomes that express the genes. Hence, you can get indirect effects as one gene being over-expressed reduces the supplies available to another gene that was meant to be constant.
The final keynote came from Chris Voigt and he showed off CELLO, the newest automated genetic design tool. The Verilog hardware description language was used to define the desired circuits function and additional constraints could then be added. This allowed flexible design inputs that use a set of defined genetic regulatory parts.
— Traci H. Angelli (@Traci_H_Angelli) July 21, 2016
Impressively, building and measuring 60 genetic circuits with different input states gave them the correct state in 92% of the 412 desired output states. This effort is a major milestone in automated genetic circuit design and the connection to exhaustive experiments was a real tour de force.
Overall, SEED 2016 showed off some of the best synthetic biology work being done outside of the cell in test tubes or in the computer. Both can be systems that help improve design and prototyping of complex synthetic gene networks.