Protein-based target recognition systems (Tyr/Ser SSRs, meganucleases, zinc finger nucleases, TALENs)

Delivering big DNA to mammalian cells

Dana Carroll, University of Utah

Leslie Mitchell, NYU Langone Medical Center

https://www.youtube.com/watch?v=PEpsXxvTL70&t=0s&list=PLHpV_30XFQ8RN0v_PIiPKnf8c_QHVztFM&index=39

I am a postdoc in NY.

You'll notice as these slides come up is that I have changed and simplified the talk title. Delivering big DNA to mammalian cells. A slightly simpler title that covers something that we have been working on in the Boeke lab. We are no longer limied to studying cells that are from natural evolution. We can build new functions and direct growth of cells iwth new function by design. When I think about this emerging field of synthetic genomics, I break it down into two big classes.

I am thinking about editing and writing genomes. This is an analogy from Joel Bader and he talked about this in the context of the Sc2 project. In some cases, we want focused edits where maybe there's a typo or word that was wrong. We can go in and make a focused edit. In other cases, there's a million edits to make, and it makes more sense to go back, design it, and write it from scratch. And then we need to deliver it to cells. For Sc2 that meant 400 bp between clusters of edits. So we built that from scratch.

So starting to think about this overview of de novo design, and how to get that from the synthesized fragments or other starting materials like BACs, PCR amplicons or other source material, you can start thinking about... this sort of layer of the workflow that involves assembly and editing, you can use cells for those activities and all sorts of editing tools to build big DNA of interest which can either be delivered to your destination cells, with conjugation by ecoli, cell fusion, or by a purified DNA that is just transfection protocol. A lot of these activities are currently underway in the Boeke lab.

For the sc2 project, the assembly and delivery were all part and parcel in a sense in that the delivery overwrote the existing DNA and homologous recombination in yeast is efficient. But we see this as a major challenge for GP-write.

Just to break down these challenges in a simple way, it comes up right in front. It comes up when purifying DNA in vitro. I hope you can see this 100 kb construct purified by celsium chloride next week, it's cloudy. It's an old school method- coming back to life. In the act of purifying your DNA, the transfer into cells is limited, and once delivered it needs to move into the nucleus, like naked dsDNA getting into the nucleus, like antiviral pathways. And then targeted delivery, and I'll talk about some of those strategies we're using.

We are in not early stages but not quite ready for prime time. So I'm going to highlight some of the great literature that is out there to give you guys an idea of the activiteis that we feel are important

Brown et al 2017 NAR, 10 kb to 1.1. MB, efficiency 1/1000

It's the idea of delivering sphere plasma mediated cell fusing, fusing it with mammalian cells and by virtue of that fusion delivering the contents into the cell. David Brown I think, I really recommend if you are interested in this, to go back and read it. It allows you to optimize based on the destination cell type. They are showing delivery efficiency independent of DNA length. The use of this sphere plasma mediated cell fusion is long standing, we have delivered megabases in the past in untargeted integration to mouse ES cells and cloned a T cell receptor which is 1.2 megabase in tact. That was untargeted-- and targeted delivery of DNA is important to think about it. Genomes can be linear DNA in the nucleus regardless of homology, so this idea of deliering, there are two approaches from two different groups

Brown et al 2016, ACS Syn Bio

Sakuma et al, 2016 Nature Protocols

You can relatively efficiently deliver a circular plasmid DNA into a crispr or TALENs DSB break.

Site-specific integratio by site-specific recombinase is another technique. It's ligation without loss of nucleotides. There are heterotypic site specific recombination systems listed here.

Stark, Microbiol Spectr 214

This is often thought of as deliery of big DNA in cassette exchange.

Wallace et al 207, Cell

Recombinase-mediated genomic replacements - they replaced mouse alphaglobin locus. They engineered a BAC to have recombinase sites, they had a selection system, engineered native locus to build a landing pad. This slide shows how they did it. It's complex. The important parts for the delivery -- landing pad, inserted half a gene, flanked, there's a lox site, heterotypic lox 511, anti-selection resin gene, on the BAC they engineered it in a way to recaptiulate the ene for.. selection... anti-selection again. Pop out that HRT gene using flip. So this is a nice strategy for the delivery of big DNA. And we're coming up with sort of takes on that approach. And the way we see this all fitting together is represented back in the initial diagram.

We want to build out in yeast, 100 to megabase size molecules in yeast, then transferred by these fusion protocols directly into mammalian systems where we have pre-engineered and characterized landing pads using dsDNA breaks. And then integrating them. And then, we're not going to unveil it, but using an approach called bigN which is a take on recombinase-mediated casette exchange to engineer the DNA into the existing landing pad. This allows editing of specific DNA and ... and then the seamless delivery without having to go in vitro to deliver big DNA into these mammalian systems.

People:

  • Laura McCulloh
  • Megan Hogan
  • Sudarshan Pinglay
  • Weimin Zhang
  • Sergei German
  • Andrew Martin
  • Henri Berger
  • Vincent German
  • Boeke lab

Q&A