Engineering intercellular communication and other complex systems: Expanding the synthetic capabilities of yeast

Virginia Cornish, Columbia University

https://www.youtube.com/watch?v=OV5QTdw5mxs&t=0s&list=PLHpV_30XFQ8RN0v_PIiPKnf8c_QHVztFM&index=24

I was originally trained in organic chemistry. I did chemical biology which combines that with molecular biology. When I started my independent career, I came back in 1990, I thought that you know there's already a lot of people doing a lot of beautiful work in chemistry. It would be fun to see if we could engineer yeast to do chemistry and that turned out to be a good decision for us.

We started with new technologies, we asked could we engineer yeast to carry out both the mutagenesis and the selection steps directly in the cell so that we could program the yeast to engineer new chemistry for us. Think about how hard it is and how library large it takes, just for an antibody, we are going to need new enabling technologies.

We started by focusing on selection in early 200s. We can combine yeast genomics and organic chemistry to trick yeast into carrying out new chemistry, like cephem hydrolysis by a cephalosprinase. K. Baker et al, V. W. Cornish, Proc. Ntat... and then D. W. Romaini, P. Peralta-Yaha, ACS Synth Biol 2012 602. This is recombination-based mutagenesis, we have a heritable plasmid that keeps track of mutations and pass on the changes. This was heritable recombinase. The kinds of library sizes for engineering multi-component systems, if we can take advantage of sexual reproduction rather than just searching a library at 108 maybe we could have a better virtual search of 1024 size which gets exciting for engineering reasons.

Really what we and others are really excited about now is that as we've had a series of enabling technologies in the field of synthetic biology is to ask the question of whether we can engineer a cell to do anything, well what would we pick? You don't want to engineer yeast to synthesize a molecule that it then leaches out of the cell; we're really interested in applications where the yeast itelf, instead of small molecule or biologic like antibody is the product. So that's really the question on our mind. Yeast makes sense- it's good in the lab and it's already a household practice and it has been safe for centuries.

The first application we thought of us was in the area of diagnostics. Basically I think the idea is very simple. If you think about antibody dipstick test, those are actually pretty expensive, they are about $1 per test and they could require expensive reagents or equipment to read them out. Could we engineer dried yeast to do the same thing? Put a modular receptor on the outside of the yeast, and then do some colorimetrics so that someone in a remote village could test by eye. Dried yeast could be about $0.01 per test and not require technical reagents or expertise.

I won't go into this project today but it's under review, it was a collaboration with Miguel Jimenez, and Nili Ostrov and Sonja Billerbeck. Under revision. Here we're sort of thinking about engineering an individual yeast as a product. One of the next things we're going to want to do is engineer communities of yeast and other organisms that can function together to carry out more complex functions. Scalable yeast communication system work has been with Jef Boeke and Agimon and Shen.

We'd like to have some communication language for cells. The best communication language out there comes from bacterial communities where they use .. lactoma.. those just aren't very scalable. They have been around for a while but we have only seen 2 to 3 orthogonal system. So we basically said hey look yeast already has an orthogonal communication language using peptides as signalling molecules. It's the sensor. So we hypothesized that we could just go in from what's known from the fugal databases and pull out the sequences of the gPCRs and the peptides from fungus organisms. We worked out a pipeline to do this, we have cloned over 30 peptide gPRPC pairs. They cluster, they are highly orthogonal, and the other 30% have a lot of cross-talk but don't cluster. So this gives us a lot of access to bigger libraries.

For the peptide module we have a framework for carrying this out. We have developed assays for seeing how it's secreted. Suffice it to say, we have 30 peptide GPRC building blocks. We're starting to play with location and show successful communication between yeast, we want to look at a daisy chain motif and other ways to do this. I think this is just the beginning. I think it will be a massively scalable communication language. GPCR.

  • Bertrand Adanve
  • Andrew Anzalone
  • Znixing Chen
  • Estefania Chavez
  • Rachel Fleisher
  • Marie Harton
  • Ethuo Herbst
  • Miguel Jimenez
  • Chaoroan Jing
  • Annie Lin
  • Andy Ng
  • Millicent Olawale
  • Dr. Sonja Billerbeck
  • Caroline Patenode
  • Dr. Casey Brown
  • mIa Shandell
  • Tracy Wang