TRE重组酶的分子进化

My name is Frank Alz. I'm a group leader at the Maxx Plunk Institute for Macular Cell Biology and Genetics and Dresden. And I'm going to talk about now our work to develop site specific recombinase for molecular surgery.

Basically as a, as a tool to be able to excise sequences, predefined sequences from the genome, and more specifically trying to eradicate HIV infected cells from the virus. So basically to excise the HIV provirus from, from the genome of infected Cells. Other conceptual Is that we adapt already well-established genetic tool, namely the Cree Recombinase to bind and recognize and, and, and use a target site which we predefine.

So we set out to say, can we make a site specific recombinase, which binds a site which naturally occurs in the LTR of an HIV virus. And if we would be able to do that, that we would predict that it should be possible then to excise the provirus From the genome. Basically The viral lifecycle goes through a stage where the, where the virus integrates into the cells and then they are through a reverse transcription, makes a DNA template, which is flanked by two terminal repeats.

They're called long terminal repeats or short LTRs. And this is then brought into the nucleus and actually integrates more or less randomly into the genome of the host genome. So now you have the situation where you have a so-called provirus, which is integrated and is flanked by two long terminal repeats by two LTRs.

The concept now was if we would be able to make a recombinase recognizes sequence that exists in the LT R, we should be able to excise the provirus as a cycle from the host genome and therefore eradicate the virus from the cells. That was the conceptual Idea. Well, the recombinase is Actually a DNA binding protein and it finds a very specific site.

So it's has a DNA binding domain, which recognizes a certain sequence within within DNA. And only when it finds this exact sequence is, is the enzyme active. It is not active on any of the other Sequences.

Principles of molecular evolution Are basically known by, by, by anyone because it's nothing else than breeding. So a principle of molecular evolution is that you introduce mutations and then you have a selection process and you select out a certain property. It's the same as when you do plant breeding.

You know you want certain, a certain size or a certain color, and you then take always again for the next generation progeny that has certain features and you start breeding them to each other. That's exactly the same that we do for molecules. So it's basically everything but on small scale.

So we start off with a, an enzyme which has a certain activity in our case that's CRE recombinase, and then we start off with a wild type protein. We then slightly mutate the site that this recombinase is binding, but it still has very, very, very low activity. Then we introduce random mutations into the enzyme, into the recombinase and we select out mutants, which now are able to bind this mutated site.

Once we have this, we can then also combine the different mutants together to finally then get to the ultimate product that we want to have so that the recombinase recognizes the sequence that we defined in The LTR. So we start off, As I said, with a random mutagen library. This is typically done by error prone PCR.

So we use PCR conditions, which whether tuck polymerase for instance, makes a lot of mistakes. And this way we generate a random library that carries mutations all over the place. We can also use a mutator strain, a bacterial mutator strain, which induces introduces many mutations in in certain transcripts.

And these are then used as a starting material. The other technology that we use is that after a while we use a DNA shuffling. That's basically a technology that allows you to combine different beneficial mutations because you don't really know which one of the mutations is beneficial, but you will have different mutants that have different mutations.

So you have one gene that has a mutation here, which is beneficial. You have a second clone which has a mutation here, and in order to bring these two together, now you perform random fragmentation of the different pieces and you then bring this together in a PCR reaction to reassemble the the products. And now you will have at certain frequency a combination of the two where part of the recombination is made out of this mutant and part is made out this mutant.

And then this guy here will have an additional effect which will improve and, and will be able, you will be able to select much faster the candidates that have desired Properties. So I would Set the definition where I say that the wild type protein has absolutely no activity on a certain site. So if we take for instance, our sequence that we selected out of the LTR out of the H-I-V-L-T-R, if we try cre recombinase on the sequence, you will have absolutely no activity.

CRE recombinase is not able to recombine the site and you have a new recombinase by definition, I would say, when you then have a, a, a enzyme which is able to bind and recombine this Sequence. We didn't Predefine how many cycles we were going to do, but we performed as many cycles as were necessary until we had an enzyme which had the activity that we wanted. So we started off with very minor mutations in the, in the site and then started to combine the different mutations together.

This was one of the important steps of why this process was working. So we went through intermediates and had to go through this many cycles to find an enzyme, which is then active enough to go to the next cycle. It's very similar to to any breeding process.

You set up a certain activity that you want, a certain feature that you want, and then you start breeding. And until you have what you wanted, you're not there and you have to do more generations. That's basically where it is.

I should also say that the 120 evolution cycles or a hundred twenty eight six to be exact, just to find a, an intermediate step where we said, so now we can stop because we have some activity. The recombination we have right now are still less active than the CRE recombinase on its wild type side. So in order to get more active recombinase, we would have to do more evolution cycles.

So it doesn't basically end Where we are right now for the evolution Process. I don't see any technical limitations. It's questionable on how far you can push this.

So the locks LTR side, so the sequence that you find in the LTR HA shares 50%sequence similarity with the natural locks P side, the natural side that career recombination recognizes. So the question is, can one goal even further, and we don't know this yet, but the fact that we only had fif have 50%sequence similarity to the wild type site and suggests that you can push This even further. The recombinase works Well in in human tissue culture cells, even though as I said, it's not quite as active as query recombinase on, its on its side.

So probably the stability is somewhat less than the query recombinase, the wild type enzyme is, and more selection would be necessary and more evolution cycles would be necessary to, to improve these features in the Future. So the first step Is that we improve the site, this specific recombinase, the three recombinase, make it more active, maybe find other sites that are very prominent sites that are, that would be good targets in an LTR and find a very good recombinase for this. We would then have to test this recombinase in different assays in first, again in tissue culture cells, but then very importantly to test this recombinase in animal models.

So we would have to check whether this recombinase has any undesirable side effects in an organism. And then of course, a very challenging aspect is to develop a good delivery system for this recombinase, which is not trivial. And is is is a whole area of research in itself to develop.

It was basically a gene therapy approach That we would have to take. Well, I think we see a Shift of medicine going more and more molecular. And for this of course, we need tools to be able to work on such a scale.

And this is where I see where the future is of the evolution process. So we need to make custom made enzymes that work on a process that we really would like to be able to use in the case of the recombinase, maybe one could take this as an analogy to, to a tool for, for a surgeon that needs to excise in an operation, something that he needs to excise. So here we need tools now where a molecular surgeon basically wants to specifically excise a certain piece of DNA from the, from the genome.

And for this, he needs the tools to be able to do this. So molecular evolution can basically fulfill this, this function. Many enzymes exist in, in nature, which are, have very desirable properties to be able to do this.

And we just have to make them to our needs that we need to, to be able to make them useful. And I think molecular evolution can play a very important role here. Where the limitations is is difficult to say at the moment because the tech, this field is, is really still in its infancy.

So time will tell in where the limitations are. I'm sure there will be some, but at the moment right now, I think there's more opportunities than really limitations.