Ultra-safe cell line

Nili Ostrov, Harvard Medical School

last year transcript: https://diyhpl.us/wiki/transcripts/hgp-write/2016-05-10/recoding-ecoli-to-57-codons/


Design, synthesis and testing toward a 57-codon genome

What if the DNA that the virus injects into the host cell could not be properly translated? We have a host cell, a wild type ecoli cell, and it has three stop codons. What if we took every place on the genome that UAG stop codon is used, and change that to a different one? If we change it to another codon, then when a virus infects a cell and uses the missing codon that we eliminated, that virus cannot reproduce. That would make the host cell virus-resistant. Does this work?

This is work from another lab using an ecoli cell where all UAG was eliminated and changed to UAA codon. Different viruses, when you test whether they can infect the cell, 5 of these 7 cannot infect the cell. Two of these can still infect the cell. Those 5 actually use the UAG codon which is missing in the host cell now. So they can't infect the cell. The other two still have other codons they use, so they can still infect.

In our lab now, after doing the first ecoli that is missing a single stop codon, we wanted to enhance this. We want all viruses to be incapable of infecting our cells. So we are recoding more than one single stop codon. So this is our current work-- recoding of 7 different codons in the ecoli genome. We're eliminating those codons from the ecoli genome by replacing them with different codons for the same amino acids.

We are not changing the proteins. We are replacing one codon with another codon. We divided the ecoli genome into pieces, and we de novo synthesized all of the ecoli genome with protein-coding genes which don't have any of these 7 codons. Just to point out the differences between those two recoded strains.... if you look at the genome size, ecoli is about 4 megabases. So if you want to remove one single stop codon, you need to change 321 positions on the genome. If you want to change 7 codons, you would have to change 62,000 codon positions. Having done those 2, our lab now has hte experience to say what does it take to do this and how to do a bigger scale recoding. This project was done by genome editing, that is-- oligos were used to change 300 positions on an existing ecoli genome. Tihs project was done on the computer, for the 62,000 changes.

Synthesis of this genome basically took a computerized designed genome and made all of these changes into it and then if you print it out and synthesize the DNA to make a very modified ecoli. And so this ecoli we expect is going to be much more virus-resistant than the previous version.

Thinking about making a human, or a CHa cell, or any other cell line, you would think: how does this scale up to other organisms? What would it take to change one single stop codon in any of those organisms? V. natriegens would be 1000 changes, S. cerevisiae would be 3000 changes, .. C elegans, D. melanogster, A. thaliana are about the same. M. musculus and H. sapiens are about the same.

When you go up in genome size, it doesn't mean that you have to synthesize more genome-- if you just look at the numbers, human genome has less codon variation than ecoli. It would be less effort to do individual edits than writing the entire genome.

Many of these edits are not related to protein-coding genes. There are many dfferent technologies we are doing to do this recoding. DNA technologies for decoding are very much available for ecoli and we're hoping to soon have the radically-recoded ecoli done. For mammalian and human cells, all of these technologies and synthesis and assembly and CRISPR recombineering are not easy to do in human cells... yet. Delivery into host cells is not easily done in mammalian and human cells.

Safe cell lines, if we can make them, are not going to be just virus resistant, but there are other interesting advantages like biocontainment and adding new chemistry by introducing non-standard amino acids.


Q: Will viruses be able to adapt?

George: I will stick my neck out and say no. Viruses will not be able to adapt that. If they have thousands of codons in their genome, the viruses are not gooing to be able to do that. Incremental change could make that a little better, like for medical drugs or something. But this is a huge change, the virus wasn't "watching" while the host was evolving. The host was able to do it while the virus wasn't watching and that's not fair.