De novo chromosomes that shuttle between yeast and mammalian systems

Alina Chan



Hi. I am at Harvard Medical. I am the postdoc working on chromosomes in the lab. Why are we interested in HACs? They have academic interest and commercial interest. We're interested in HACs because you can hack into chromosomes not just humans many different things. Maybe lizards. This is the goal, just HACs chromosome biology.

So what makes a good HAC? In terms of a synthetic genome, what do we do with a HAC? There's a way to score HACs for us. Segregation -- centromere competence. If you can't segregate your HAC, it doesn't work. The next thing is compatibility. This is what HACs leverage over random pieces of DNA, this is compatibility with the system. Yeast work with these, they express recognizer, in human biology we haven't breached that barrier much... The next thing is organization. We don't just create a jumble of DNA, you want to know what's in your HAC, what it looks like while it's in the cell, and does it get passed on with integrity to the next cell. It must be replication competent. And also, expression is super important, if you can't express off your HAC then it's not useful. I think we've met significant progress on these issues. The next big one though is shuttle-ability, being able to shuttle yeast vectors into bacteria and manipulate and construct them and then put them into mammalian cells. Megabases of DNA in yeast and shuttling into human cells would be hugely beneficial compared to our current methods.

So the last two decades of HACs... one way is top-down, so taking .. a human chromosome, and what he did is he chopped this up, removed the bits between the centromeres, just the core and the end, about 5 megabases have been truncated, inserted into chicken cells. Taking small piece of centromere repeat sample, amplifying it, throwing it into human cells and hopefully getting something that persists as a HAC. This is what Vladimir's group has done.

These hACs now go into cells where you can load genes into them. And then use microcells to carry the chromosomes. We're trying to go back to square one and make more well defined HACs or make it a more efficient process.

Our modest beginnings of making HACs, this is what we've done, I'm going to run through it. We've redesigned the hAC and centromere and we also made a way to asseemble it and delver really large megabase HACs. Also, design of centromeres, which are really repetitive. This is a problem with recombination and rearrangement. We went back to this consensus sequence from 1987 and we removed as much of the repeat as possible, keeping only conserved motifs. Many of the sites can be varied. The longest stretch of any repetition is really just that lack of repeat. We make this so that the LacI repeat ... many different proteins into human cell to help centromere function.

This is our HAC design. We wanted to be able to shuttle between bacteria, yeast and human cells, and we want to build it in yeast because it's good at assembling long and complex sequences. The HAC can be linearized by cutting. So you can have the optio nof making your HAC linear. We also have selectable markers like GFP and drug resistance and also it's an origin of replication. Also a centromere which has a way of self-perpetuation. Also human telomeres. We're in process of putting in the telomeres, but we're testing it with a centromere in it. This is about 30 kb in size, which is really shocking. So a centromere as big as 5 kb in human cells is capable of being a centromere.

The way we done it through assembly took about a year because of all the repeats. The best way is to do tire assembly, through it all into yeast, and this works better than any in vitro method out there right now (Gibson assembly).

Also, we have delivered megabases of DNA from yeast into human cells. This is a really big deal. You can now create HACs of megabases in yeast and then push it into human cells. You can do this and have 1 in 300 cells get the hAC. If you have yeast that have the HAC, you can deliver proteins. My collaborator has been able to give herpes and delivered this to human cells and shown that the human cells get herpes.

So right now, I'm going o run through it a little, the key ingredient is arresting the cells in mitosis, and this removes the nuclear membrane and htis makes it easier for the HAC to get into the cell. One megabase HAC, there's virtually no difference, about 1 in 300 or 1 in 1000, which is pretty amazing.

That concludes my talk. We're testing these HACs now. This has been funded by DARPA. Also collaboration with Pam Silver, Jeff Way, John Glass, David Brown (JCVI).

Q: That was a marvelously clear presentation including the slides. That was great. I wanted to ask, which different types of mammalian cells are you doing?

A: We have protocols for yeast mammalian infusions, even mosquito cells, but first we're doing HAC cells. I don't expect it to be an overnight miracle. It will need some work.

Q: .. highly competent DNA.. integrated... did you observe.. into chromosome?

A: We haven't gotten that far yet. Up to today, .. it's been a week since then. I haven't been able to check if it's integrated into chromosome yet. I have only checked propagation. You don't want your HACs to be randomly inserting, that's a great comment.

Q: Did you say there's a 5 kb centromere that's functional?

A: 5-10 kb of repeats, thrown into different cells, and put them through....

Q: So multimerization of 5 kb?

A: This is a huge question about whehter you get repeats or not; as far as I know, they just inserted, into an existing chromsome, and they showed that it forms at that site.