DeSciNYC

DeSciNYC

← Back to videos

Talk & transcript

Genome Editing in Early Human Embryos

Dr. Diego Marin

Transcript

Download text

Automatically generated transcript. May contain errors. Select a timestamp to watch that moment on YouTube.

0:00

[Music] Test test. [Music] Test one, two, three. Glad everyone is here today. Thank you all for coming to the first ever DIN NYC event. Okay, it's not true. >> Okay, so we have a lot to cover. So, we are going to And if you're in the back, just come come forward a little bit. You'll um you'll enjoy the the being able to hear. And also like a pro tip um

0:48

along the edge is often time you can hear people really you can hear all of this space you can hear really well and like this part by the um the stairwell you can't hear well but this part over here you can hear well but you just have to go around the back. So that's sort of just like a a little bit of a pro tip. If you can't hear it's it's because you're in a bad spot. Okay so we will jump right in. So the first thing we want to start with is um the DIN NYC. So what is DIN NYC? It's like a science meetup in New York City where you can meet and chill with other people who are also interested in science and the

1:28

future and embryos and all that sort of thing. So if you're here, it's probably because you're interested in science. And there's too many run clubs and too many drinking clubs and too many bowling clubs in New York, but not enough science. So this club is the the change to make that happen. So we've had so many speakers. This is actually like our 35th one doing this. So that's kind of crazy. Like there's a lot of We've done this many many many times and we all the videos are all online. Um so the next one we're actually going to do one in two weeks.

2:02

We it's a surprise speaker but mark your calendars. And the one after that is going to be on nutrient transportation and cell metabolism. So cell metabolism and and the circadian rhythm are all super super interesting and becoming more and more important as we learn about longevity and we learn about sleep and we learn about, you know, how to be healthy. So we're trying to focus on that. And also if you guys have ideas for speakers, um, come up and chat with me after because we're always looking for people that are cut doing cutting edge stuff. um uh around science.

2:38

Um so um what is decentralized science? A lot of people are wondering, you know, I like science, but this is called DINYC. Why why is that? So science is undergoing a lot of changes right now. And if you've looked at like the the news, you'll see like a lot of PE funding is being taken out from the government. The government is no longer funding a lot of the research. And some people say this is very very bad. But this actually creates an opportunity for science to happen at the periphery and for people to fund science on their own.

3:09

And this is what decentralized science is about is about uh citizen science and allowing people to choose what types of things get funded, what type where their data belongs, um who they want to invest in. And I think for too long, uh science has sort of been trapped in the ivory tower and not been able to be able to be seen by other people. Um, so this is like a a new opportunity for for a new wave of science to happen. So I'm going to chat around um what types of things are going to happen with this. Um so the first type of thing that we can uh think about for decentralized science is we're going to do a conduct a sleep study. So

3:49

we're going to look to see how sleep is impacting people. and sleep. Uh if you guys know Brian Johnson, he says the first thing you have to do is you have to understand uh you have to get your sleep in order and then after that you have to get your um that'll help you get your uh food in or your exercise in order and then after that you sort of uh get your um your uh food your your spoot sleep exercise food and um if you do that you have all sorts of cognitive enhancements. you can live longer, things like that.

4:25

Um, old slides. So, sleep is really interesting and the there's two examples of people that are doing interesting sleep research in a new type of way. So, there who how many people here know what what sleep 8 is? So, it's this bed that cools itself off automatically, which is very interesting. And during the open AI drama when Sam Alman was ousted from from open from open AAI, the founder of sleep 8 said, "Are people sleeping better or worse? Are they is there a lot of anxiety happening that's causing them to like be awake more?" And what he did is he was able to look at the data from all these different beds and find that

5:07

people on average slept there's a 27% increase in people getting under 5 hours of sleep during this open AI drama. And what's so interesting about this is this this wasn't a research study that was sort of conducted in a university or that required like intense approval or anything like that. This is just something that people were automatically like giving their data and allowed the crowd and it allowed a new a new type of person to conduct research. Um in a similar vein there's this like sleep is exercise sleep and exercise. So I've seen a lot of people here wearing a whoop and uh the whoop is really interesting. So what someone did what the this founder of Whoop did is he

5:47

recently published a paper in Nature where they're looking at Whoop data and how exercise impacts sleep depending on how early or late people do the exercise. So if you if you exercise closer to bedtime, they found that people actually get worse sleep. But this is also another example of a type of research that is done in a in a different type of way than sort of in a traditional academic setting.

6:12

So, what we're doing is we're going to create a uh we've created like this sleep lab that a lot of people are participating already in New York. We have uh over 10,000 people sleeping each night uh using this app that we were about to describe and uh it's very cool. So, it works with any wearable device and we're trying to create like an open sleep research platform where people can contribute their sleep data and then anyone can ask questions of their sleep data as well.

6:40

Um, so we have different point systems. Uh, you can track your patterns. Um, and we're already collecting interesting data around that. We're finding that people when they eat closer to their bedtime, uh, they actually like get a significantly worse score. Um, if you're interested in trying the app out, you can download it here. Uh, if you download it and you show me after, I will give you this beautiful sleep mask so you can participate. If you're interested in sleep research, um we're also hiring, so we really need some um more developers and business people. Um we're funded by Polychain and a bunch of other VCs to try and distri make science distributed. Uh so if

7:21

you're interested in in working with us, that is also a possibility as well. All right. So is there any questions? >> Okay. So Oh, yeah. Go ahead. >> You said 1.2 out of how many >> out of 100? So it's an interesting problem of how do you like determine a score for sleep, right? So like what is perfect sleep? Some people might be optimizing for recovery. Some people might be optimizing for longevity. Some people might be optimizing for uh like uh feeling having good mental clarity clarity. But we basically created a normalized sleep score around uh REM, deep sleep, light sleep and then created

8:02

a 0 to 100 sleep score to be able to predict how about how well people slept. All right. So without further ado, we will go to our main speaker of the night. So human embryo editing is the future. So in the previous talks we actually we had a speaker who was talking about embryo selection where they had an embryo and they were doing uh genetic analysis on it and then finding sort of the healthiest embryo to implant. But this is taking it sort of a step further where um uh now you can actually before implantation you can edit the egg and uh you know select for any characteristics

8:44

that you want. So, this is going to be fascinating. And we have Diego from Genomic Prediction, which is a new company, which I'm sure he'll tell more about. So, without further ado, let's give him a round of applause. Thank you very much, guys. >> Okay. So, thank you. >> I do one picture. >> All right. Three, two, one. Three, two, one. Well, thank you. Good evening for the introduction and for coming here. I know

9:25

it's late at night, but um this is a topic that we have been working for a while. >> Oh, okay. Sorry. I had a shot here. >> Oh, like the speaker. Can we put them down so they can see the slides? >> Okay, >> we're good there. Everyone can see. Perfect. So yes, I probably you saw that tonight Stean Jedc was supposed to give this talk. However, he had some issues

10:08

with some data that he's going like this. Okay, so it's first time I give a talk like that, but it's fine. So he was go he's going to give a talk here, I think probably this summer, if not in this fall about the continuation of this project. But today I would like to share with you a story um that was a collaboration between our company genomic prediction. I'll tell you a little bit of what we do and deer Egley and his lab here in Columbia University.

10:37

And I wanted to start this um slide with this video. I don't know if you guys have seen this before, but this is a human embryo being cultured in an IVF lab. You know, we have several clinics in Manhattan, in the city, also in Jersey, around the world. And basically this is when the first day of development. This is a time lapse of 5 days of development. Usually we culture these embryos until day five/6 depending some ambros are slower and faster than others. And um it is until we get the day five where we get the structure called the blastosis. And this is going to be important for this talk. So I just

11:18

want to put those background so that we can understand what we're doing with gene editing in the brain plantation embryo. So these blastoes there are two major components. We have this bunch of cells inside which are called the inner cell mass. They give rise to the actual fetus to the baby. Whereas that coat of cells around it would be what we call tacctom and that's what gives rise to the placenta. It is important because all the testing that we do is from the tropha. Okay? So we don't touch the actual cells that become the actual fetus.

11:54

Now again I want to put this give you the setting of where this magic happens and that is in the embryology lab. So the embryology lab they are different from clinic to clinic but basically this is what once looks like. This is one in Panama, one of the best um of quality in the world. And you will see this micromanipulators of the microscope. All this imaging that I'm going to show is happening in a petra dish. Right? So sometimes it's just hard to visualize how this is taken. So imagine that all this is happening here with what we call a micromanipulator. All those needles that you'll be seeing in the videos is like a jostick. You play with it and then you have a pedal to move it. But

12:35

these are things that sometimes are not shown and I just wanted to clarify the setting of where this magic happens. Now before we start I do want to clarify two different concepts. One is genetic selection and one is gene editing which is the topic of today's talk. The reason why I say it is because sometimes I hear these terms interchangeable and they are completely different. Genetic selection is something that we're doing every single day in IVF clinics that do PGT or genetic testing. Now, before we go to gene editing, I did want to go over briefly what genetic testing means.

13:16

You're going to listen to me saying PGT a lot. So, PGT would be pre-implantation genetic test. That's what our company genomic prediction did. So, in pre-implantation genetic testing, of course, you need the embryo. So, this embryos comes from the clinic. Again, you put the em the egg with the sperm, fertilize it, let it culture, and then once you have that blastoise, that embryo that I showed you at the beginning, then you're going to take a biopsy. I'm going to show you a video shortly about how that looks. That biopsy is sent to the lab. We amplify the DNA. We interpret the data and then we just give a PDF to the clinic and the patient so that they can select which embryo would be transferred into the

13:56

womb. Now this is a video of a biopsy again just for you guys to see how this happens. We have this embryo. The shell is called zona palucida. Some cells are coming out. We take that biopsy from those cells coming out that would eventually remember become the placenta. And with a laser it's like a pedal. So it's actually like a nice instrument. And once you do the laser you kind of suck it. Then those cells are put in a PCR tube and sent to the laboratories.

14:26

What happens to this embryo? It goes to liquid nitrogen and it could be there for years, right? We have pregnancies I think after 20 years of an embryo being biopsy and tested. So nothing like we were talking before about turnaround times of genetic testing. Liquid nitrogen that's not really an issue. Okay. Now that we clarify that concept, gene editing is completely different because in gene editing you're actually messing with the sequence of DNA. So you're making changes. Again, genetic selection, you're just screening, seeing what the embryo already has. With gene editing, we are actually changing the DNA of the those cells.

15:10

Now you might have heard the term crisper cast 9. Um for gene editing, there are several tools. This is not the only one. However, this is the one that's been more developed. Uh there was a Nobel Prize to two people who developed this crisper cast 9. I won't get into detail, but it's a system that already bacteria has to code against you know foreign invasions of virus or other um um kind of organisms. But basically the crisper cast 9 became so useful because once you understand the components and you get them into the cell it's actually pretty effective in doing this job. So crisper cast 9 has four major components. One component

15:52

would be the cast 9. The cast 9 protein is this Mickey here. So it's an enzyme. It's a protein that goes to the DNA and cuts it. That's component number one. The second component is what we call the guide RNA. So the guide RNA is a molecule that uh you can actually design the sequence whatever you want. You send there is a company called IDT. So you send them the sequence what nucleotides you want and they will send you a tube with um some powder which is liophilized primer. You put some water you have it there. It's very important because then you can design whatever sequence you can target it to any location in the genome.

16:33

The third component is of course the target DNA and in our lecture we're going to focus on the DNA inside those embryo cells. And the fourth component is something that gets ignored when we talk about crisper cast 9 but it's extremely important and I try to show it here which is the cell DNA repair mechanisms because without this crisper cast 9 is just useless. So I'm just going to take it out because for purpose of showing how this works. So how does gene editing works? First of all, we have and I won't go in detail into that but there is a whole research on how do we get these components into the cell.

17:11

You could get them of course by injection which is very easy but it's very hard to inject you know liquid into a single cell which is more than microscopic. You could do electroporation many other techniques but let's just pretend we're already inside the cell. So we design a guide RNA to a specific sequence and what will happen is that guide RNA would hybridize would match to the DNA. So it would kind of open it. This is happening in the cell. It puts the whole machinery there and the cut happens.

17:46

It's literally a cut. So it generates what's called the double strand DNA break. And this is not editing. This is just cutting the DNA. The editing happens when the cell DNA repair mechanisms go and fix this because when that's fixed, it's not perfect. You introduce novel nucleotides insertions or you can delete nucleotides, deletions or you can actually let's say there was a T in that position the cell could just put an A.

18:17

So that is how the gene editing actually happen. It's not that you are going and customize and put an A there. you are relying on the cell repair mechanisms and therefore as of now it's somewhat random. There are some technologies that are trying to be more precise but this is the concept of what happens. Now in 2017 18 there was a major publication of our field from the group of Dr. Malib in Oregon and this was the first proof of concept in the human embryo of a successful gene editing. So basically what they did they took the egg they fertilized it with the sperm then they

18:58

let it culture one day. This is the stage of a zygote. You have the two genomes mom and dad. And at that stage they injected the cocktail of proteins the components that I showed you before. Now I'm going to show you a video that has nothing to do with this paper. It's from our laboratory. But just so you know how these injections happen. So basically you have it there. You load your needle. you put it into the embryo.

19:24

That's how it's happening. Okay? And then you let the embryo culture, right? This embryo was cultured until day three of development when you have around eight cells. We call them blastoirs. And then you analyze them. So the experimental design was pretty elegant in this study because what they did was to find a sperm from a human who was already affected with a recessive mutation. So all the DNA in that sperm had a mutation, right? Therefore the only outcome in that embryo was to be heterosyg um you know for sure because the sperm already had it meaning that if this

20:06

actually was working then we would expect some cells to be fixed or not to have that mutation. So this is what they did exactly like that uh mutant sperm put it in the egg and then they got several embryos of course I don't remember how many but one of the outcomes was what was expected which is all those cells would have the mutation so their crisper cast 9 did not work but then in some cells they started seeing that some cells actually didn't have the mutation at all and most importantly few of those embryos not many because You can see it's a mixture of things. Again, it's very random, but in some of those embryos, all the cells

20:48

did not have a mutation. So, you would understand the impact that this had in our field because for the first time, we were fixing or curing a mutation in the embryo before it's transferred into the womb. The problem here and what happens in science and something that I have learned in my years of this uh field is that you have to be very careful over interterpreting data. The analysis that was done in those cells, those blastoirs, those embryo cells was pretty shallow analysis. And just because you didn't see the mutation, the assumption was that you cured it, right? It was

21:29

also in the news and derley which is Stefan Jarave's mentor and with whom we collaborate started writing these papers trying to provide different interpretations to the same finding. So why hold on maybe you are not curing the mutation. Maybe you are just not seeing it. Maybe that chromosome was not even there. And of course unfortunately in science there is a lot of egos, right? And then you start seeing all these back and forth back and forth. So even the authors of that initial paper they um brought up this article saying that this was yeah it made a lot of sense but was

22:11

sheer speculation and it was true. It was just speculation and if you want to prove a concept you just have to get it done. So what we decided to do with ether and genome prediction was to replicate the experiments you know add reproducibility because reproducibility is something that it's essential in science especially when you're doing a biotech company or something like that you need to have this reproduced in a different setting by different people with different interests.

22:39

So basically we follow a very similar experimental design. We took a sperm from a guy who was blind. So he had both of his copies of that gene had the mutation. They were affected. The gene was EYS. So the guy was blind. He donated the sperm. We took donated eggs and we fertilize those eggs with the sperm from the blind um guy. And then the experiment was to once we bought the embryos inject the crisper cast 9 directed to that EYS gene make blastosis. you know the day 5 a.m.

23:15

development at this point do PGT which is taking some cells from the Tactom to make sure that the editing happened right so that was the first proof of concept and another thing we wanted to do and we're still doing is because this embryo is not going to get transferred you know we can talk about that like in the future when are we going to transfer editing embryos but this embryos we took the inner cell mass which develops into the fetus and developed a embryionic stem cell because when you develop an embryionic stem cell you don't have a hundred cells to play with you have hundreds of thousands millions of cells and this is important because what if this crisper cast 9 is cutting somewhere

23:57

else we are directing it to cut to one gene but it could certainly cut somewhere else and that's what we call offtarget effects and until we have safety data I do not believe it would be a transfer of embryo into a patient so anyway Now one thing that was a gamecher for this study remember that I told you before that there was some suspicion that there was over interpretation of data. So how do you analyze that cell to make sure that you are actually not seeing that mutation and you are not causing any other harm. There are several ways to analyze embry

24:38

several technologies. You could do fish, you could do something called array genome um arbitization, you could do next generation sequencing. But there is a specific platform called snip array. So a snip array is a platform that we use genomic prediction and it has it genotypes up to 800,000 locations in the genome. And that these snips are small locations that are polymorphic. So maybe some of you would have an A alil, I'll have a B al. So they are highly polymorphic. In fact, you can actually predict ancestry based on just these locations. So anyway, long story short, we had access to this platform and we just decided to use it on those cells

25:19

just to make sure whether the editing was happening correctly or not. Now before I show you the results, I do want to walk you through how we read this data. So the company that we work at it's it does PGT. So it's a commercial company. We have we processed last year 40,000 embryos and this is a small company like these are embryos being processed every year. So the plot that we get is called a violin plot. So remember that we have 23 pairs of chromosomes. Well 22 and then the sexual chromosomes. So here each chromosome is represented by one of these diamonds. And what this means

26:00

basically I don't want to go in detail but as you could see the y-axis has a little frequency. So if we have two copies of each chromosome we would expect the vast majority of snips of these locations that are hetererozygote meaning differing a little from mom and dad to be clustered around 50/50 because we have two copies of the chromosome. So that's why this diamond kind of localizes 2.5 because this is confirming that this embryo has two copies for each specific chromosome except for chromosome 16. Again it's just an example for chromosome 16. You can see that there is no data in the middle and what this suggests is something in genetics that we call loss

26:42

of heterosygosity. That is because there's only one copy of chromosome 16 and this is therefore an abnormal embryo. In fact, we have so much resolution um per embryo. Remember this is coming from just six cells of that are being tested. You can visualize the entire chromosome short arm and long arm and you can see that there is no data in the middle. So this is how we do confirmations that there is a loss of a chromosome. And of course the report of that patient would say you have an abnormal embryo with a loss of chromosome 16. Don't even waste money and time transferring it because it will result into an abortion or you know a biochemical pregnancy. Anyway, the

27:24

reason why I show you all this is because this tool was not used before in the previous research and as you could see it gives you a lot of information on these cells. So now I'm going to show you the results. So just to recap the experimental design, we took the egg um that was donated, the sperm from the guy who was blind, fertilized the embryo, injected the crisper castine cocktail, developed the embryo until day three and then until day five. We took the biopsy for PGT and tested. Remember that that that gene that we had for this specific uh experiment was in chromosome 6. The gene was called EYS and is located here

28:07

close to the centromeir which is that part that unites both short and long arm of the chromosome. And this is one example that came in the paper. So as you could see this is for those who work in the IVF industry. This is a very beautiful embryo. It's very good morphology. So that was good because we were not messing up with the morphology of embryo development. And once we visualize all the chromosomes, but this is just the example of chromosome 6, short arm, long arm, we started seeing something that was unexpected. If you look at the short arm, remember what I showed you before here, you see data around 0.5, which is

28:47

expected because you expect two copies of each chromosome. But what was very unusual is that right at the point of cut of crisper cast 9 we started seeing loss of hetereroygos. We thought at the beginning that it was a denovo um what we call deletion because this means that you only have one copy of that part of the chromosome because it happens every every day like you see the noble duplications or deletions but it kept happening and always right at the point of cut that we designed the crisper castine to go in other embryos. The finding was even more striking. We were seeing loss of

29:28

hetereroygosity everywhere in the chromosome. And after doing so much data and data and seeing the same thing, we concluded that crisper cast 9 unlike what they interpreted in the previous paper, it's not that it was getting rid of the mutation. It was getting rid of the entire chromosome or part of the chromosome. So that's why crisper cast 9 was actually not fixing the mutation. it was eliminating the entire chromosome or part of it.

30:00

So you can imagine the implications of this. You know this was an unexpected finding. We presented in several conferences. There was a lot of I think I have a couple. Well, this is the paper that came out from Michael Zukado. It actually helped us to understand a lot of cell repair mechanisms in the embryo. But the major finding was that crisper cast 9 was not doing what it was intended to do. it was actually eliminating DNA.

30:28

Now the cool thing again I always emphasize on reproducibility after some groups saw these findings they replicated the Francis Creek Institute in London which is very has a lot of prestige um also in the European meeting there is also other groups using this nipper and they're finding exactly the same. Now again you would expect every when things go okay people are like okay when things go wrong everyone is on India right so if you could see all these uh titles of all these major um newspapers crisper is bad crisper is not ready crisper will never be the thing the solution for gene editing and

31:11

it's true because we found something that wasn't expected and remember we need safety data before we do this in the embryo But there is something in the field of infertility that we have to cope every single day and that is embryionic annuploy. So embionic annuploy first of all is the most common genetic abnormality in humans. What does aniploy mean? Remember that we we as humans we have 46 chromosomes 22 pairs plus a sex chromosome.

31:42

If a human or an embryo has more chromosomes than 46 or less than 46, that's considered an anuplo. And what's even worse, at least for our field, is that annuploy is the most common genetic abnormality of infertility. So this is a plot that we did when I was training at RM of New Jersey, a huge IVF program in Jersey. So we took patients from all maternal age and as you could see the older the woman gets or the egg gets the higher the rate of embryionic annuploying because remember that women are born with all the eggs that are going to be produced during her lifetime. Right? So a woman

32:23

who is 40 or 45 years old has eggs that have been there since she was born. So the aging of X results in having embryos with annuploites more extra or less chromosomes. But the reason why I bring this here is because for infertility this is a huge issue. That's why infertility clinics we get a lot of older women because they try and try and try and reason why they don't get pregnant most of the cases is because their embryos have an extra chromosome for example. So it was actually sitting down and be okay we have potentially a solution to eliminate chromosomes which are already ahead for the infertility clinics and that was um

33:05

something that we started doing. So again I won't get into detail but this was the first time that we could find a tool that could get rid of embryionic annuploy. So cure tricoms for example cure down syndrome from the um pre-implantation embryo. So the experimental design basically it was to take the egg remember that an egg is produced through a process called gamtogenesis and then you have a stem cell that divides into three two cells and then these two cells into other two and there are two cells that are a byproduct of this division which are called the polar bodies right and the polar bodies are these small cells that

33:47

get DNA because it's a division so they get equal part of DNA but they just don't get the cytoplasm and the rest of chemicals only one cell becomes the egg. But the reason why it becomes more important is if we can actually analyze the polar body and let's say one polar body has an extra chromosome 16, we can assume that the egg has the missing chromosome 16 or vice versa. If the polar body has a missing chromosome 16, we can assume that the egg has the extra chromosome 16. So the experimental design was to do this and then if we find that the polar body has a missing chromosome 16 why don't we inject crisper cast 9 directed to that

34:28

chromosome 16 so we can get rid of it and this was actually performed here in Columbia University so Jenna Turosi who was the fellow by then she started doing these experiments again you will analyze the polar body actually I show you the biopsy of the embryo but back in the day before we didn't know whether we could touch the embryo. We were doing polar body biopsies, right? That was the old standard before. Now, no one does it because the egg is extremely delicate and most of them die. But in this case, we managed I mean this were for research. These are all human eggs. We took the polar body, we analyzed it and some of those polar bodies, at least for

35:08

this egg, one has amazing chromosome 13, but the other one was very useful because it didn't have any signal of chromosome 16. So we could assume that the extra chromosome was in the egg and if that egg was going to get fertilized that will be an embryo with three copies of chromosome 16 the two maternal and the paternal one. So after we did that we designed the guide RNA remember this is the part of the crisper cast 9 you can design directed to the chromosome 16 from the mom because you didn't want to get rid of the paternal chromosome 16.

35:42

It's important that you have the two copies of each chromosome, one from mom, one from dad. Not in all, but in most chromosomes. So anyway, what happened is that we did this injection. We cultured this embryo until day three. We took each cell and analyzed it. And we had we got actually a myriad of results. Some of them had three copies of chromosome 16. So we confirmed that our assumption was correct. Some cell had that exchromosome.

36:10

Some of them duplications, deletions right at that point of cut that we designed the crisper cast 9. But what became more striking and surprising was that some of those cells have only two copies of chromosome 16. And this was the first time that has been reported that anuploic correction can happen in the human embry. Now this of course opened the door to many interviews, many things. Um there is it was actually awarded a prize for the ASRM. The ASRM is our society in the US. It's called the American Society for Reproductive Medicine. Jenna Turus got the first award because this was something

36:52

extremely unique that no one had reported before. Now, of course, we're very far away from applying this clinically, but the research continues and this is already I mean, we have some reviews I think in nature communications, but right now it's in bioarchchive if you are more interested in seeing the detail and the minutiae of this project and um well then the press started to get more positive. So, okay, Crisper is not that bad. Maybe Crisper could have another application. In fact, people are like trying to get into it and there is a lot of other reports not in human embryos but they're doing in cell lines and it's like oh these guys are right this is possible. So again

37:34

reproducibility and hopefully in the future we could have some solution to do this for people who are 45 years old their eggs are all abnormal and this is the only chance to have a baby without recurring to gamut donation. Okay, now this is just I'm almost done. This is just an other um let's say application of this specific for very very rare cases but you know we all have the chance we should have the chance to have a baby of our own genetically. So there is something called a Robertsonian transllocation. So this is just a patient um a carotype that we generate.

38:14

You know when you want a carotype you just want to know how many chromosomes you have and for chromosome 21 there are some individuals that do have two copies of chromosome 21 however they are attached to each other that's what we call the robberian transllocation why is this a problem those people are normal you know phenotypically the problem is that when they make sperm or when they make eggs then they are going to have half of those sperm and eggs with um newly 21 meaning like no chromosome 21 because the two copies are attached and the other half of the sperm or the eggs are going to be having two copies because they are attached. So if you fertilize an egg with that sperm is

38:56

going to be down syndrome no matter what or vice versa. So imagine if we could have something like this. We have this guy with this problem, that condition again, you know, typically normal. We get an egg. We get this embryo. But what if we can get rid of that extra maternal DNA? Now, of course, the baby would have the same robberian transllocation and we can discuss ethics about it. But at least this would be a solution for that person, that couple to have a baby that would be phenotypically healthy. Okay.

39:31

Okay, so this is basically what I had for you guys. This is a story that continues. This we started this collaboration in 2021. Now there are so many questions that has been have arise after this project for example once when you inject this Christopher castine cocktail I mean it remains there you cannot take it out so how do you make sure that it stops working like you don't want it to keep editing. So Alex Robles he also presented in ASRM 2022.

39:57

We were trying to do a second injection which is an inhibitor of crisper cast 9. So we could stop tell him okay do your job at the beginning when you are one cell but stop when you are two or three more cells because you don't have to have you don't want to have mixtures right of these carotypes. So this is something that's already in bioarchchive also under review.

40:16

Um Ian Rasheep is also a fellow from Columbia University and uh then we went to another chromosome which is problematic which is chromosome 18 also presented as 2023. Uh last year we presented some data that um is very enticing and very cool because I'm talking about eliminating chromosomes and employed but remember that initially we wanted to cure a mutation right and that's a project that keeps going and hopefully Steph and Jereek would tell you a lot of good data of how we found a solution to do this accurately.

40:52

Okay. So, and I just wanted to end this talk with this quote. Nathan Trebis, he's my mentor, also the co-founder where I worked and I really like it because I do agree with it. I don't know if you can read it and maybe discuss about it, but the future of IVF, gene editing and preventation genetics is precisely curing disease before pregnancy. This is what we should aim, what we should go um after it. We were talking before about companies offering um IQ testing or skin color, cosmetic traits.

41:27

We can debate that, but at least curing disease before pregnancy should be the way to go. And with this, I of course want to thank inside New York. Um but most importantly, well, of course, my lab, my team, we have a lab in Hacketttown, New Jersey. And most importantly, where the magic really happened, which was Derek's laboratory. This is when they came to genomic prediction to pay us a visit. This is a steepen that he will come probably later. Marcus Kman and of course university. So here is my contact and hopefully you enjoyed the talk. Thank you.

42:00

[Applause] >> Does anyone have any questions? >> Yeah, go ahead. >> Do you want to come up and then being live streamed also? Uh just here I'll >> take the mic. Okay. >> Um yeah, you you mentioned this like goal um at the end of like curing all these disorders. Do you think people who are thinking about having kids should like wait um you know is this like an imminent development? Are you confident your timelines at all?

42:42

So when it comes to gene editing per se, I don't think we should wait because it will take time until this will be available to transfer an embryo. However, there are other things that are available that is not edited. For example, the embryo selection. So for that that's available. But for gene editing, maybe I think five years will be you know underestimation for gene editing. But for example, selection of polygenic traits or avoid some mutation like cystic fibrosis. That's something that is being done nowadays.

43:23

>> So if I understand correctly, uh Christopher cast 9 doesn't do precise editing, but it can knock off entire chromosomes. Um, and I think in your research you showed that, you know, you were able to use crypto to knock off the chromosome, but not all the cells get the chromosome knockout, right? So, in that case, is it ready for uh implantation or do you still have work to do because not all the cells have the same?

43:50

Yes, of course. There is a lot of work to do because we just tried it in chromosome um was it I think 18 and six and some of targeting phase we don't see it everywhere in the in other chromosomes but we have to try them on all of them right so that was one and then you mentioned something at the beginning that I was going to answer it >> it didn't work for all of them right >> oh yes so exactly so again when do you inject it because if you inject that crisper castine cocktail after you already injected the sperm you're going to have something called mosism which is like once imagine like I mean it's very hard to control but what we're trying to

44:30

do is injecting this cocktail at the time where you inject the sperm so when you inject the sperm you kind of load it in that micromanipulator and you can put the needle on like a drop with all these cocktails which sucks it and if you inject it at that time then you have less chance of having mois in different cells with different you Garex, but um that's something that we're still researching.

45:01

>> I wanted to ask you about the scientists in China that years ago Chriser basically corrected or introduced a mutation twins, human twins uh to make them uh resistant to HIV infection. um the twins were born and they were you know they growing up now. So I was wondering if that was luck or they were very good in selecting the embryos making sure that the procedure didn't eat chromosomes.

45:39

So probably some of you have heard about that Chinese scientist who did this some years ago and um he claimed in a conference that um he did this Christopher cast 9 transfer the embers and indeed these babies were born resistance against HIV. The problem with this is the secrecy behind it. Right? So there is the even the Chinese government got involved. So this guy was in jail and he's no longer in jail. He's actually having his research lab. So the secrecy behind it has kind of blinded us to know exactly what's happened there.

46:13

The babies and I think the couple are also like under some kind of protection. But this is something that was such a huge hit for the scientific community that um they got kind of isolated you know after he went in jail and unfortunately we don't have those reports and this is very important because what I'm showing you here and what you will read is what's being reported in literature right there is I mean imagine all these crazy people in like these isolated islands you don't know what they're doing we were talking before about Mexico in Mexico there are all these laboratories doing all sorts of things right that we don't No, but to answer your question, there is a lot of secrecy behind it, but the guy claims to do it. It's just that we cannot even

46:53

sequence it. I wish to sequence the genome to see what we find. >> I would say it's probable that if they didn't result in live births for that type of thing. >> What? >> No, they like the likelihood that that resulted in like a live birth, I think, is near zero. >> Yeah. Right. That's what they That's what he claims. >> Yeah. Yeah. So, I mean, it's like the guy who claimed he invented Bitcoin, but then like doesn't back it up with any proof is like it's probably just like riding off of like the hope that it it could be true, but I don't think it actually worked.

47:24

>> Right.

47:35

>> Uh yeah, really interesting stuff. So um for the anniplated correction this is happening before it's an embryo right first of all okay um and what what I was wondering was um it do we know why the crisper cast 9 takes out an entire chromosome rather than replaces a part of it or um and then also Um, does it only take out one of them and leave the other when you do it in an annual platey?

48:15

So, um, regarding the first question, so we do these injections once you already have an embryo because you already have the sperm inside. So, that the embryo with the two circles, that means it's already what we call a zygote or embryo day one. And we want to do it there because if you go beyond then you're going to have mixtures. But yet when you have the day one of an embryo that's number one. Now why is crisper castine cutting and getting rid of all these embryos? Oh sorry of the chromosomes.

48:44

There are several reasons for this. Um what we found is that remember like I showed you you have a chromosome and you have what's called the centromeir which is that part that unites both arms. We found that the closest you cut you cut to the centromeir, the higher the rate of eliminating chromosomes. And that is because these chromosomes, they use the centromeir to attach what's called the spindle. The spindle is some sort of a bridge that the chromosomes locate there to separate for cell division. So what we believe is happening is that once you cut there, maybe the DNA mechanism is is not perfect. Actually, we found out that it's very not perfect in the human embryo. And because of this lack of

49:26

repair, we believe that that little cut could cause the whole loss of that chromosome. Right? So, and there was the third one that I forgot. >> So, in an annual does it take copies? >> Yes. So, in a remember that when we wanted to cure the anuploid or get rid of that extra chromosome, we needed to be very careful because we didn't want to get rid of the two normal ones or we didn't want to get rid of the paternal.

49:52

That's why of all these embryos and sorry sperm and eggs that we have for this research we also got DNA from the from the gamut provider from the guy or the woman because then you have to sequence it so you can be okay this chromosome has this snips or this sequence which is specific from the egg and this is the one we want to cut. So we designed a guide RNA specific for that extra chromosome that was from the mom. Exactly. Because you don't want to get rid of the paternal one or the other normal one. Why don't you remove the chromosomes at the egg stage? Why?

50:24

>> Because at the egg stage, in order to know whether there is an extra chromosome or a missing chromosome, you have to have some sort of information. So that's why you have to do the polar body, right? And the polar body gets extruded once you get the sperm inside, right? So once you get actually an egg, it's so cool because the egg hasn't actually um done meiosis or become an egg until the sperm gets in. Once the sper gets in, there's a phenomenon called egg activation and then it extrudes that second polar body and it becomes an actual egg. However, at that time it's already an embryo.

51:04

Um, do you still get the umal deletion from single stranded breaks like using primary for example? So I won't answer that because that's a Stean Jeravex talk. So yes remember something very important crisper cast 9 is a huge let's say I always imagine there is an umbrella you have crisper cast 9 and you have all these different systems crisper cast 9 base editing crisper cast 9 double strand brakes or there are a myriad of crisper cast 9 the one that I'm presenting here is the more canonical version that cuts both strands of DNA but if you manage to cut maybe

51:47

one strand maybe you won't see that result but hopefully that will be you know the product for Stephan stock. >> What are the ethical concerns of this science and like where do you see it converging? >> So ethics is very hard to give an answer because what I've learned is that ethics is a dynamic concept. It depends on the society. I'm from Costa Rica originally. I can tell you that the ethics behind this is very different from the US and even Europe. I mean I would what I could answer is mostly based on safety. I believe that if we were to do this study here we will need a lot of or safety data until we can actually transfer it.

52:27

But ethics is something that we all will have our own opinion you know and depending of our background also. >> Say it again. Say it to the mic. >> What diseases can you cure today? Okay, if you mean cure with actual editing, if you think of it any disease that's has a single mutation because for example, we have several things. Let's say cystic fibrosis, that's a small mutation, one letter in the nucle in the sequence that will be easy to do. In fact, we have some embryos that have been uh cured. I don't want to use that term, but that's something. Fragile X

53:09

and Huntington disease. These are repeats, expansions. That's something that we don't even start thinking of how to do it. So I would think that everything that's a single mutation could potentially be fixed. But again potentially because another assumption we make is that the whole genome is the same and there are some places in the genome that are so hard to get in so hard to cut for some specific epigenetic things. But as of now I would say like if it's a specific mutation that's pretty pro possible to cure or to edit.

53:40

Do you have an example? >> Cystic fibrosis is one for example. Uh betatalismia I believe it's another one. I mean there is a myriad of of of these diseases but point mutations I would say. >> What are your predictions for the next five years of you know advancements you expect in this field? >> So in the gene editing field I think the predictions is we're going to have more offtarget effects anal an analysis.

54:09

remember what we do in these embryos. We don't throw them away. We create embriionic stem cell lines because I mean it we have to to make sure that it's not damaging other parts. So I think that's going to be a huge area of research. Maybe AI would give us a tool to try to cope with this because it's just a lot of data and a lot of position. So that's something that I actually do not do not think there is a biotech company or a niche to explore this but this is for sure something that will be pivotal before we actually transfer an edited embry.

54:42

How far do you think you can get with like the technology that we currently have? Like we we currently have crisper cast 9. We we there there's certain things we can do with it and not all of them are researched. How far can we get with that? And what and so how far can we get with that? And what what is like we need something else? >> So when you say how far you mean in terms of what we could do in the genome or >> Yeah. What we do? Well, I mean, how far I was a fan of Jurassic Park, you know, when I was growing up and it it sounds silly, but there is a company called Colossal in Texas doing things very similar to what they're doing. Listen, the genome is something that as long as

55:23

you understand the raw material, it's it's it's something that it's very easy to do like all these sequences that you hear, you just send them to just do a and in silicon model you send the order to the lab and next day you have the primer. So how far it's hard to say but I mean you could edit I don't know one thing for example people ask me like what okay what do you okay cure disease whatever what else can you do there is something called protective varants there are some humans for example that have mutations that don't cause disease but could be protect protective against Alzheimer so imagine someone who has family history of Alzheimer and you can use gene editing to actually not fix a

56:03

mutation but introduce a mutation that's protected against alzimer So I think that's something that not many people have discussed but to my knowledge I think it has a lot of potential [Music] >> this process shows you come up come up so here >> um so my question is how you how would you edit the genome for a human that's already alive as opposed to editing like the the origin cell of of an embryo and then having that problem.

56:39

>> Yeah. So this is a I mean there is a whole area of research of gene therapies using percast 9. Now of course this is another scope because here the point is just to have all cells in our body with the same editing but for example I think for cle cell anemia what that some groups are doing is that they extract cells from someone who is affected with some disease. they use crisper just to fix that population of cells to make them let's say wild type or not have the cle cell mutation you reinject them or you put them back into the blood of that person and then I think there are very good results so in terms of editing the adult I think will be very much localized to a specific type of tissue

57:21

specific type of cells but this is happening and these are the gene um therapies for cle cell for example Uh so when people are thinking about doing invivo gene editing, it's like the the edit on an adult is like obviously like mosaic, right? In fact, you want it to be mosaic because you want to target like specific tissue types and stuff like being mosaic is considered a success, right? Like you want to edit every cell. Um, and so like given that we're so like fine with having like mosaic adults, why is mosaicism considered so like dangerous and something to avoid in human drumline editing?

58:00

>> I think the major reason for that is that those cells at that early stage they are there are different stages but they are called plur potent multi-otent. So for example one of those I mean that inner cell mass which is what becomes the fetus it's just like a 100 cells and that would get you know develop into all the tissues and organs that we have. So I think the reason why mosism is such a problem in the embryo on the early embryo is because you don't know what cell what's the destiny of that cell what if it's going to be a brain what if it's going to be just the the the skin so it's because at that time it has the potential to develop every tissue in the

58:40

body and that's why it become so hard because you don't know the fate of it >> one thing I didn't understand was there there's like a one in4 chance it successfully works and the rust fell when you do the crisper editing. So yes, this is certainly not 100% because remember we are relying on the cell reaper DNA cell reaper mechanisms. So part of a step and stock is like some of the efficiency we have got with this editing around 80%. But as of now we don't get there and the reason is mostly

59:20

because it's some of an stoastic approach you know we depend on what the cell is doing. So maybe there is another niche to do some research what what does the cell need to do their its job well right and that's something that we haven't focused but I believe that's something that could justify how inefficient it is as of now. >> So what's the probability now that it works? This depends a lot on different settings. At least in our project with Steen, we have around 80% editing, successful editing on all sales, but 80% is still something that would be clinically not acceptable, right? But we'll have to get maybe 99. I don't know.

59:57

>> The very first paper was like 10% >> the metallic. Yeah, but again, it was not actual editing. Then we discover a misinterpretation. But yeah, it depends also on the Chris N. So we have go to her first. Yeah. Um, so you kind of briefly mentioned that some of your competitors are focusing more on these more controversial behavioral traits, IQ, that sort of thing. Do you think that that's like leaping ahead too early in the field and field really needs to focus on disease right now or is that interesting science to explore whether or not these polygenic risk scores for

1:00:37

behavioral traits have value? Yeah. >> So this is a I will answer it. It's just nothing to do with gene editing. This is embryo selection. So there are some companies who are doing this PGT testing and trying not only to predict risk of disease but also the predict the IQ of that embryo or cosmetic traits. Um now my opinion again this my opinion I believe that as of now we should focus on clinically relevant disease because you know we are still these are patients after all right and we don't want to do harm but then you start with the ethical question of reproductive autonomy this is something that in the US is is is you

1:01:20

know liberty right something very important in Europe is something that not even think about this so that's why when it comes to ethics this depend depends on the context. We get a lot of patients from East Asia and they just want the perfect baby. You know, they want the most intelligent. So, it's something that yeah, in our company, we don't offer, but we get requests all over the place. Is it right or wrong?

1:01:41

I'm not in the place to answer it, but it's certainly possible. That that I can assure you it's certainly possible and technically accurate. Also, >> did anyone hasn't asked a question yet? Want to ask ask one? >> Oh, yeah. So I'm just wondering rather than going through the extremes of gene editing or deletion in crisper is there a way to more um I guess naturally stimulate gene expression whether in um you know younger chromosomes that are just expressing themselves or adults to kind of optimize how they express themselves because you know we all have same 23 chromosomes. It's just how they are expressed that we get certain diseases

1:02:22

or certain um benefits from it. Um >> well yes so that's so the question is is very interesting because without you know messing with everything like is there some sort of environment maybe that we can create that would decrease the the level of the disease. This is something that in our field of the IVF industry has been of course a holy grail like we all want to go there. We want minimal invasive procedures. The thing is that as of now nothing has worked. So the pre-implantation embryo is it is delicate but it is it's very strong like

1:03:03

those blastos is you know you can make them in an incubator even with mold like they they are pretty strong. Sometimes we believe that we were talking about that before like nature is perfect. I disagree. I think we as human beings homo sapiens we need to understand nature but there is nothing wrong with going beyond it or trying to perfect it at least to avoid disease in the in the embryos but to answer your question there is a lot of of approaches incubators with different gases with different environments but as of now nothing that you can see an actual cause or effect >> one more >> for the anoploy correction. What is the

1:03:45

road map for that? What's possible from things that viable >> bonding? Well, several things. Fonding I mean of course it's an idea that we have of course presented it several times. There is a patent behind it. Um I think part of the issue is that it's a very complex thing to implement. It sounds pretty cool, but you have to do a polar body biopsy and you have to go through that. Nowadays, women who are 42 years old, they still make embryos that are uploid, which means no anloid. And we already have a screening tool. So, we already have an screening tool that is $200 per embryo. So, this would be

1:04:25

specific for specific cases. And I think that's one of the push backs, but the research is there and there's a lot of people interested. One more thing >> how's it going? Um, so question is I guess if this is going to be responsible for like increasing IQ going back to that will have like cascading effects down the line see like 300 IQ. So um with this IQ it's first of all it's is it's the the heritability of IQ meaning like the background the genetic background is pretty strong. That's why it's so accurate in prediction. Now

1:05:05

remember that when you talk about selecting for IQ, you are doing a screening. It's not that you're editing and making the 300 IQ. You are just selecting from what you has. Actually, we get sometimes question that we want like this high IQ. Well, first of all, we don't offer it. Second of it, we can only select from what you're offering, right? Like we that's true. Sometimes you want the tallest and we are sure like that that cannot be done. But I would say that I mean a lot of people are interested here. You have no idea the amount of emails that we get from all over the world. Do we do this? Do we do this? Is there going to be potential gene editing to increase IQ in the future? Listen, is it technically possible? I don't see why not. But is it

1:05:48

going to be implemented? Then that's another question. But it's mostly screening right now, right? Of a pool of embryos that you generated. >> That I don't know. One more question here and then we'll do personal and he'll be around after. >> What can we do like daytoday like in our daily lives to get like better sperm better embryo like ourselves with our team? So to improve the quality of gamuts, there is a whole industry of this you know I think that your question came um I mean I don't that's not my expertise but certainly if if you follow our industry the IVF there is and there's a bunch of companies trying to what can

1:06:28

you do to improve the quality of the sperm and eggs we were talking about egg freezing before that's I mean they asked me for my advice I think beyond 35 that should be and you don't have kids that you would like in the future to freeze your with the sperm after 40. I think that would be also a good thing to do, but that's another industry that I don't really cover.

1:06:51

>> Two more. Go ahead. Go ahead. >> Um, we're we were talking about this a little before, but what's like the best way to mitigate against like um potential unintended consequences of like optimizing for certain genes and like what's what's out there right now? Yeah. So we were discussing at the beginning of this talk that there is a hu a very legitimate criticism against selecting for these things because what if you are selecting for I don't know the embryo with the high the lowest risk of developing type 1 diabetes and at the same time you are deselecting for another trait.

1:07:29

It's of course very hard to know. However all these predictors are validated in bioanks. Bio banks remember are this data of different countries with hundreds of thousands of individuals who are already probably dead. So we know whether they develop a phenotype or not they have a disease or not. And all these data suggests that once you do simulations of the genetic selection you are not really seeing the prevalence of an extra disease. And that's actually what opened the door to offer these predictors because if we would have found like oh okay in these patients with type 1 diabetes you are actually I don't know going down in creativity I don't know then that will be a stop but as of now there is no data

1:08:10

suggesting that it could be true however as of now you know we don't see that we actually see positive pleotropy which is when you reduce the risk of type 1 diabetes you reduce the risk of other disease because of coorbidity which makes sense but other traits we we still don't Cool. All right. Well, thank you so much for you have one question or >> you want to ask did you think of using the repair mechanism that exist in the cell because the cell has a

1:08:53

disease. It remembers Yes. >> So again, crisper >> crisper cast 9 depends solely on the or not solely but depends a lot on the cell the cell repair mechanisms. But something that we found in this we already knew in our field but something we found in this project is that the human embryo is very inefficient in repair because maybe it doesn't have that much memory there. Right? when you create these gamuts they go through an epigenetic deletion or eraser. So we believe that's why it's so poor in repairing and that's what we have anuploit is that what that's the most

1:09:33

common genetic abnormality otherwise we wouldn't see it but that's the material we have with and again answering your question before maybe we can inject something there in those embryos that will just enhance these mechanisms of repair and hopefully that combination could be super efficient for this edit >> amazing well let's give our >> if you have questions he'll be around after so feel free more. Thank you guys for coming.

1:10:15

>> Hey, how's it going? Michael. >> Oh, perfect.

1:10:30

[Music] [Music] Oh, perfect. I never [Music] down. [Music] Happy birthday. [Music] [Music]

1:12:02

[Music] Looks like [Music] [Music] in my team. [Music] Lional. [Music] [Music] There's a lot of good

1:13:40

actually [Music] very [Music] You're having [Music]

1:14:11

[Music] Okay. [Music] [Music]

1:14:42

Nice. [Music] Happy [Music]

1:15:12

[Music] dance. [Music] with that

1:15:45

[Music] It's all [Music]

1:16:17

standing.

1:16:26

[Music] I was thinking the moment I was over the minute I was thinking the moment I was [Music] thinking the moment I [Music] Open up the morning slow down. [Music] Oh my [Music] go [Music] The moment I was over the minute I was thinking the moment I wasing [Music]