Podcast - The CRISPR children, episode 4

Alison van Eenennaam of the University of California, Davis talks about Cosmo and germline gene-editing of cattle. And she considers the children with genomes edited before their birth. (Hint: She thinks that experiment was a terrible idea.)
Podcast - The CRISPR children, episode 4

To accompany my story The CRISPR Children in Nature Biotechnology, I am producing a rolling series of podcasts. This is episode 4 with Dr. Alison van Eenennaam of the University of California, Davis 

You can listen to the podcast here. 

You can also find this podcast on Apple podcastsGoogle podcastsSpotify and wherever else you get your podcasts. It's part of a series called Conversations with scientists.

Here are the other episodes, each with a transcript.

Here is episode 1 of The CRISPR Children with Kiran Musunuru from the University of Pennsylvania, also with a transcript 

 Here is episode 2 with Rudolf Jaenisch, Whitehead Institute

Here is episode 3 with Eben Kirksey, Deakin University

And here is a transcript of episode 4, a chat with Dr. Alison van Eenennaam of the University of California, Davis

Note: These podcasts are produced to be heard. If you can, please tune in. Transcripts are generated using speech recognition software and there’s a human editor. But a transcript may contain errors. Please check the corresponding audio before quoting.

Alison van Eenennaam

Sorry about that. I was spinning plasma. Sometimes the red blood cells are like “yah, I don’t really want to separate from plasma.” So you have to spin it for a bit longer. I knew if I had an appointment. It was actually Cosmo’s plasma so you can blame it on him.


Hi and welcome to Conversations with scientists, I’m Vivien Marx. That’s Dr Alison van Eenennaam, a researcher at the University of California Davis talking about Cosmo. And you will hear more about him and Dr. van Eenenaam shortly. I asked her to pronounce her name so I could try to do it.  

Alison van Eeenennaam

van Eenennaam


Could be Dutch?

Alison van Eenennaam

It’s Dutch, it’s my married name. I’m Australian, my husband is Dutch a few generations back. Dutch dairy farmers as fate would have it. 


Before we get to Cosmo let me just say a few things about this podcast. You can find some of my work in Nature journals, which publish studies by working scientists. And they publish science journalism by science journalists like me. This podcast episode is a podcast about farm animals and it’s about human babies. And germline gene-editing.  

Some cautions for you. If you don’t like meat, you might not like this podcast. Although you might want to hear about projects related to livestock health and breeding in the tropics or about reviving and restoring endangered species.

 You might not like this podcast if you do not want to hear about animal experiments. Although we do also know that many things intended for use in people are tested in animals first. And that is indeed fraught.

Even if you have aversions of this kind, I would like to invite you to stay tuned to hear more about Alison van Eenennaam’s work.

I asked her about her research and about Cosmo and about gene-editing and about her thoughts on the fact that a team of scientists have gene-edited the genome in human embryos, implanted them and the pregnancy was taken to term. And I talked with her about the fact that some scientists think the technique of gene-editing is ready to make what are sometimes, and unfortunately called designer babies. Babies with genes edited before their birth.

As you may remember in 2018, He Jiankui and his lab in Southern University of Science and Technology in Shenzen, China took it upon themselves to gene-edit embryos and implant them into the uteruses of two women from whom they had removed eggs. These eggs had been fertilized in the lab.

As I produce this podcast He Jiankui is soon to be released from prison. The three children that resulted from the experiments in his lab are, according to sources I cannot name, doing ok. But that is very hard to know for sure.

Not just because there is so much secrecy about them. But also because the gene-editing didn’t, it seems, quite work in their case.

He was trying to knock out a gene, the CCR-5 gene. This would in theory make the individuals later resistant to HIV, the virus that causes AIDS.

According to his presentation at the international gene-editing summit in November 2018, Nana, one of the babies, now a toddler, had changes in the CCR5 gene that might make her resistant to HIV. But her twin sister Lulu had a deletion in only one allele, one of the two copies of the gene. So she might not be HIV resistant.

There is a third child, whom I call Amy who resulted from work in this lab. Amy’s genome is also apparently only edited in one of the two CCR5 alleles. All of these children are likely genetically mosaic. Not all of their cells have edited genomes.

 I wrote an article about these children for Nature Biotechnology and the link is in the transcript and I did a few other podcasts on these children. This is episode 4 of the series The CRISPR Children. There is no sequence you are welcome to listen to them in any order. And there are transcripts for each of them.

 Gene-editing can in certain cases perhaps help to avoid a heritable incurable disease. But as of right now scientists tell me gene-editing is not precise enough for such measures. And of course there is an ethics discussion to be had about whether such germline gene-editing is advisable or permissible. Right now it is neither.

But there are some scientists, He Jiankui is one and there may be more, who have decided the technology can already be applied to human embryos. And that these embryos can be implanted in the body of a woman and be taken to term.

But as of right now, it’s generally not considered permissible to do such experiments. Of course, policies and practices vary from one country to another. And yes there are clinics one can find that offer the service of designer babies to parents-to-be.

I asked Alison van Eenennaam what she thought about this kind of offering. She is a mother of three with two living children.  

Alison van Eenennam [5:20]

I’ve got two pretty good kids, I’m pretty happy with them , pretty happy with what I got the nature way.


And you wouldn’t if you were approached with a brochure  gene-editing.

 Alison van Eenennaam

I would laugh because the whole notion is ridiculous. The only reason ever would be a homozygous lethal condition that there’s no other way around. But even then. We’re not at that level of precision. I’ve lost a baby I have a still born and I wouldn’t wish that on anybody.

Let alone 10 recipient women , the whole emotional toll of that. It’s one thing when it’s a surrogate cow, but when you’re talking about a baby and having a loss because editing didn’t go correctly, that’s emotional toll right there.


The emotional toll of gene editing gone awry with a potential baby, would be enormous.

 So if heritable gene-editing were ever to be adapted for use in people to avoid a deadly, incurable condition for example, that cannot be avoided by genetic screening, before any of that is even considered, one would need to know more about how gene editing can be made much more precise than it is now and to understand, among many other questions, how this gene-editing might affect development.

Perhaps there are lessons to be learned from work with farm animals. They have a longer gestation time, than, say, mice do.

Cattle are not a model organism or research organism as they are sometimes called, such as fruit flies, which are heavily studied.

Meat and milk producers want to know how the genomes of their animals can be altered, to for instance produce more milk or to have higher meat yield. They can do so with selection as they breed these animals. But they can at least in theory use gene-editing. They certainly are keen on genetic knowledge. There is for example the 1000 Bulls Project. Here’s Alison van Eenennaam

 Alison van Eenennaam [7:40]

The history of that is maybe six or seven years ago when 1,000 was an audacious goal. They go together a group to do it. There was much contribution from companies, such as artificial insemination companies wanted their bulls done. They put their data in there, Hans runs the analysis, Kristy does. They make a variant calling file available to everyone in the consortium. Members benefit. It says this is where the SNPs are in the bovine genome basically.

Anyone can join. It used to be 30 animals. We got it, we did the seuqwencing the Nature biotech paper, we had enough to get in. That gave Titus and other at UC Davis, we are one of the members. I haven’t sequenced enough, they update the file every year , you needed more bulls to get into the next. I am not in it at the moment. 


The 1000 Bulls Project has identified over 100 million genetic variants.

Alison van Eenennaam [9:35]

The reason there’s so much data is because there’s a commercial interest in getting the genotypes, holy grail of finding functional SNPs. That is the rationale. I am not really  there’s only a handful of genes that have big effects on these quantitative traits. They tend to be polygenic, there’s no magic bullet. I laugh when people say they’re going to make smart kids. As if there’s an edit that makes you smart, these polygenic traits, and especially in animals.

Plants have resistance genes resistant to rust. Much more polygenic in animals. Most of our traits are quantitative and are best addressed by selection. Phenotypes and genomic relationship matrices, which is not as sexy as a single gene giving milk yield. There are a couple of startling examples, there’s one gene increases milk yield, protein. There are a couple of startling examples on gene dgat one variant increases milk yield and the other increases protein concentration. It has this really big effect. Dairy producers select for it. Because it has this big effect. And handled differently normal quantitative effect, big effect  gene and you take note of the genotype there.


Variation matters for cattle breeding. And in case the industry ever decides to present gene-edited cattle for beef production.

Alison van Eenennaam  [11:48]

It’s important to understand that variation, as we approach the FDA with our animals. They want us to show there are no unintended alterations. And every animal has got millions. I don’t’ know which cone I did and which one God did. It’s not a static genome. In the Nature Biotech paper we had recently, we looked at genome edited bulls. There’s no more variation than you see between Holsteins and Herefords for example, in fact there‘s more variation that between our genome edited bulls. To me it’s a little bit of an unanswerable question to prove you didn’t alter anything, when there so much background alterations.  


Now it’s time for Cosmo.

Alison van Eenennaam

He was April the 7.


Cosmo was born on April 7, 2020. Alison van Eeenennaam and her PhD student Joseph van Owen did gene-editing. They used CRISPR Cas9 in bovine embryos to insert—using microinjection--the SRY gene. They inserted this gene into a region on chrososome 17 called H11. That’s a so-called safe harbor region where no essential genes would be disrupted.

They also ferried in the gene that encodes for green fluorescent protein, GFP for short. Their idea was to use the GFP as a reporter gene as a marker to help with screening to identify the embryos with the inserted SRY gene.

What’s special about SRY, the sex-determining region Y, is that it encodes for the gene that shapes male sexual development in cattle They wanted to insert this gene. That takes cutting the genome and the cell repairs the cut and includes the inserted gene in the process.

They also wanted a specific kind of the repair mechanism: they wanted it to be homology-mediated end-joining, which it seems more efficient than the classic way cells repair DNA breaks and that is called homology-directed repair HDR.  So they wanted homology-mediated end-joining, HMEJ. And they wanted all this to take place on the X chromosome.

So they did the editing. Nine embryos were transferred to surrogate cows, one of which became pregnant. Nine months later a male bull calf was born. Cosmo. He weighed 110 pounds. He is the world’s first bull that had a knocked in gene that was mediated by HMEJ. Here’s Alison van Eenennaam.

Alison van Eenennaam [14:35]

What was unique about him. Mostly knockins are done in somatic cells, that’s when the HDR

pathway is active. In embryonic cells that’s not a very active pathway . it’s relatively easy to get knock outs but not easy to get knock ins.

 He’s the first bovine with a large knock-in. There was the intentional knock in which was the SRY gene. But to help us in the situation with large animals; it’s super expensive to do embryo transfers, so I can’t throw in 20 embryos if only 40% of them. are edited. That’s 20 cows, x keeping them for a year is expensive.

 We need a way to identify which animals had the knock in, which embryos. we added

GFP marker gene so we could screen for green eggs and ham basically, right, so we had a way to screen. So we had a marker, that was how we were able to successfully got a knock in.

You can get higher rates of knockin in with some of the newer techniques but it’s hard to get bigger templates in. And that was what was kind of unique about him. As it happens it went in with multiple copies. It will be inherited I think as a single/ It’s basically one locus, we’ll be able to clearly see the hypothesis works.

If SRY goes to an XX individual what’s the phenotype. If the phenotype is an animal  that doesn’t grow as well as a regular male. Then it’s not of interest. We don’t know. We have to test.  

Sometime you get normally appearing males, they’re infertile. You get ovi-testes. They are ambiguous in their appearance. We would hope that we get animals that look male

they’ll be infertile for sure but they need to grow like a male.  Otherwise they wont’ be useful for food production, beef.  


Cosmo’s genome was edited before his birth. And he has a knock in. Deletions are more common with gene-editing but in his case it’s a knock in. He is the first bull that has a gene knock in.

Gene editing uses the cell’s ability to repair such breaks, which happen all the time. DNA has elaborate DNA repair mechanisms. One of them is called HR for short homologous recombination. HR is not or not very active in embryos. Alison van Eenennaam considered a different approach, HMEJ or homology-mediated end joining. Which she had heard works better for knocking in genes.

A lot of work in gene-editing is done in cells and in mice. Alison van Eenennaam follows closely what happens in mouse labs. What as she calls them ‘the rodent people’ do.

Not only did she want to knock in a gene, she wanted a way to see, if the embryo had been gene-edited or not. So they added to the sequence that they wanted to edit in, a so called GFP construct, a section of DNA that encodes for green fluorescent protein. That means if the gene was knocked in you could check and see if the cells were edited. Expose them to a certain wavelength of light and the protein will be activated and glow.

Mind you, this is not something that would be permissible with human embryos. And just a reminder here, editing human embryos and transferring them and taking them to term is not permissible in human embryos. Alison van Eenennaam is working with cattle and she was going to try something that had been done in mice.

Alison van Eenennaam [18:35]

Again the rodent people have been working on trying to get increased homology directed repair in embryos. There was a 2017 paper in mice, where they added. Normally you have a construct, arms, they added the CRISPR target site on each end. And in that paper they had 40% knock in. Joey said let’s try that , we’d been having trouble. We used that and we say we were getting around 40% knockins with that approach. 

 Since, then. You see the idea, you try it, you transfer, there’s a pregnancy. In the meantime the mice people have tried other approaches, using single stranded DNA template–I think it’s called Crispr easy--you have the targets on the ends but you use short homology arms like 50-100 basepairs and you can get a big gene, I think maybe the biggest was like 1.3 kb, which is for a reporter gene. But SRY in our case is 1.8. No one else has a good approach especially when we added GFP in there, then it became 4.5. That’s why we chose HMEJ. And we were getting 40% then we just had to screen them, pick the green ones and transfer those to the cows.


One of the issues with gene-editing of embryos, for example cattle embryos but this is also true if you edit the genomes of human embryos, is that you don’t know if the gene-editing has worked.

In the case of the gene-edited children, what seems to have happened, is that the gene-editing didn’t quite work. In all likelihood the children are genetically mosaic.

Some cells are edited and some are not. And there may have been so-called off-target edits, those are unintended genetic changes in different parts of the genome.

You can’t draw direct parallels from animals to people or from one animal species to another. But there are lessons to be had from animals. Here is one. Sometimes the backbone, the template used to ferry in the gene to be knocked in, also gets inserted into the genome. That happened with Cosmo. Alison van Eeenennaam talks about this template integration and work with mice in which that also happened and about mosaicism.


Alison van Eenennaam [21:35]

I think there’s a lot of mouse literature. Mouse people do this a lot. Jax Labs makes knockins for us for various experiments, we’re doing. We’ll routinely get 20 different founders 10 will get thrown out because they’ve got backbone. 5 will get thrown out because they integrated at some random site and then 2 or 3 with tandem insertions at the target site, then you get one that has the gene targeted correctly at the target site. It happens so routinely that nobody thinks anything of it.

We spent a lot of time figuring out when exactly to introduce the editing reagents to reduce mosaicism. We spent two years going into M2 oocytes, just prior to fertilization. We were trying to target the X chromosome originally for SRY. We wanted the X to have the SRY and then the Y would have its own SRY and so all offspring would be male.

The X chromosome of the male is going to be present in the oocyte, the y is gonna come from dad and the x, so x is going to come. We were trying to target the X chromosome of M2 oocytes, which means we in vitro matured them  then stripped the cumulus cells off them , microinjected them, then we put the cumulus cells back on them, then we fertilized them, cultured them for 7 days. Then they went to blastocysts. Doing all of that really induces the probability of them going to a 7-day blastocyst, which is what we transfer to surrogate cows.

We through if we could get that knock in, not even at the zygotic stage but rather at the oocyte there’s only one X chromosome , there’s no opportunity for mosaicism if we did that.

We did that for a paper in Scientific Reports. What we found if we did it 6 hours post insemination: so in other words mature, fertilize, let them sit with the semen for six hours  then strip the cumulus and the semen away, then do microinjection—it’s called denuding when you do that—we got better blastocyst rates and no difference in the knock-in rate or the mosaicism rate.

It’s a tradeoff, the longer you leave the semen in there, the better fertility you will get, and the more mosaicism you will get, if you go in later.

Ideally, you go in early but that decreases your fertility and you will get less blastocysts. It’s a numbers game. We’re typically working with 400 blastocysts, if you go much below 10% you’ve only got 40. After you microinject them –they don’t appreciate getting micro-injected–so then you’ve got maybe 10 to transfer and some don’t look very good. Ideally, you would like a nice selection of healthy zygotes or blastocysts to transfer.

It’s always a numbers game. Microinjection takes a long time, too, and the way it works for the 6-hour fertilization and micro-injection. Typically we start microinjecting at around midnight. If you do too many, you get too far after fertilization. After about 4am you just lose it, you just can’t do a good job anymore. There’s a finite limit in terms of the personnel too. We spent a lot of time optimizing all that. We found the six hours.

I have no idea what the human person did. But I understand those children were mosaic, so I’m kinda guessing that he probably went in later rather than earlier, that would be my guess. But the paper hasn’t been published anywhere has it?


The paper with details about what  He Jiankui and his team performed the gene-eding in three human embryos has not been published. Since it’s considered unethical work, it’s not clear how one might publish it for example in order to have the scientific community study what has happened in the case of the three toddlers who were gene-edited before their birth.

In my story in Nature Biotechnology and in other podcast episodes I ask researchers about this and how one can learn from what he did. There are three girls, Lulu, Nana and a third girl I call Amy and who has different parents than the twins Lulu and Nana. And they are likely genetically mosaic, some of their cells have been edited and some have not. It’s unclear if they run a particular risk because of that or because of the gene-editing more generally

Alison van Eenennaam does not work in human embryos. She works with cattle. She found a way as she describes to avoid Cosmo being mosaic.


How do you take care of mosaic people

 Alison van Eenennaam

I don’t know


Indeed it’s unclear what kind of medical tests Lulu, Nana and Amy might need because of the gene-editing that was performed on them. But what is highly likely is that they are genetically mosaic. And it also appears that some cells are edited and some cells are unchanged, which geneticists call ‘wild type.’

With farm animals such as cattle and sheep, it’s not considered useful to have mosaic animals. Alison van Eenennam explains. 

Alison van Eenennaam [28:20]

That’s the worst of all worlds because then you have some wild type left In our case. I have animals on the ground and we are looking at inactivating proteins to get a certain phenotype. We have mosaics, and full knockouts, bi allelic nobody’s left in the shop. The mosaics are effectively useless. If they don’t have the phenotype, I don’t know if it’s because they are 60% wildtype.

 And if they do have the phenotype, I’ll never really know if that’s the full expressed phenotype or if that wild type proteins is modulating the characteristic in some way. With these sheep it’s a high-growth phenotype. We have those two mosaic animals out there. We did that using 18 hour microinjection that was before we had been doing all the work on the earlier .

This is how the history happened. Historically in AI, artificial insemination for dairy cows, and it’s done everywhere. You collect oocytes, you add fertilization media, around 4pm, there is an 18 hour fertilization, everyone went home, they came in in the morning and hat was then about 18 hours


These are eggs in any phase getting fertilized?

Alison van Eenennaam [29:30]

They’re zygotes getting fertilized. They’ve had culture for 24 days, to get them in vitro matured and then they’re  fertilized for 18 hours, you get good fertility, you get lots of zygotes they get transferred and everybody’s happy. That’s the protocol everybody used forever. Is that really the best approach for what we’re trying to do ? Or when it’s a one-cell zygote before it splits into the two cell stage to try to reduce mosaicism .

That’s when we started: what if we fertilize only for six hours. Well our fertility goes down but we get reduced rates of mosaicism. So for us that’s more important. If instead of getting 40 blastocysts we get 10 blastocysts but they’re all non-mosaic, that’s a better bet for us then getting 40 mosaic blastocysts. Those are the types of things in a long-lived species like cattle where you have three years to get to maturity and have their own offspring, you don’t really want a mosaic because you need to breed that out of them. And that’s a decade.

Typically they do that in mice all the time. They’re often mosaic, in fact the way they do their editing it’s almost invariably mosaic, they take them out at the two-cell stage. They’re typically not doing in in vitro. They typically let the mouse mate with a male and then they flush that zygote out.  Usually by then it’s at the two cell stage. What we do in cattle is a bit different to what they do in mice.

They don’t care that much about mosaicism, they can breed it out in 3 months, they just need a transmitter that has to go into the germline. If the rest of the somatic body is , they just want germline founders, if they get that they’re happy. They typically don’t pay any attention to the founder line because it’s known to be mosaic and there for unreliable. When I order a from Jax labs, they won’t give me the founder, they’ give me the F1, then they know at least that it’s heterozygote, hopefully homozygote in a perfect world and every cell has the same genotype, which is the definition of a mosaic is that not every cell has the same genotype. 


The idea with gene-editing in animals such as cattle is to explore new approaches to cattle breeding and also to explore basic biology. It’s also a way to learn more from mammals with a longer gestation time than mice.

For others, this work might offer insight—down the line—how one might or definitely might not apply germline gene-editing to humans. The case of germline gene editing, the offerings of designer babies definitely are ethically dubious and scientifically it over promises.

 Alison van Eenennaam

I can’t think of a single trait, beyond a recessive lethal condition, in terms of designer babies, I‘m not sure what traits we have that level of understanding of. Maybe eye color. Maybe

Even in livestock I can come up with a handful of genes that you might edit. But I certainly  can’t come up with 100 even with all our sequence data. These traits are not single gene traits, they’re not Mendelian. The only Mendelian traits are recessive autosomal


The idea He Jiankui had was to confer immunity to HIV by editing the CCR-5 gene. And one wonders if some might think perhaps one could edit the ACE-2 receptor. That’s the receptor that SARS-CoV-2 latches onto on cells. This is the virus that causes COVID-19. Might someone try to edit in immunity to COVID-19. Lots of issues there. And it’s likely SARS-CoV-2 can find other ways into the cell. Alison van Eenennaam finds this concept of editing the ACE-2 gene to confer immunity scientifically questionable.

Alison van Eenennaam

You can vaccinate them, too.


Yes, now we can.

Alison van Eenennaam [34:10]

I must admit, I haven’t followed what he did. What we do in livestock is so fundamentally different to me in terms of ethical issues. To conflate the two is a slippery pathway in my opinion. We’re growing these animals to kill and eat. That is their function in the world. The minute you start anthropomorphizing. How long is Cosmo gona live. Until he produces semen. So I will never get longevity information on him because we don’t keep farm animals for the term of their natural lives. There’s fundamental differences in terms of the role in food animals versus human editing.


Beyond the role of food animals, there is the possibility of using cattle to explore various types of questions in reproductive medicine. Just in terms of approaches, human reproductive medicine labs use similar equipment to a lab like the lab of Alison van Eenennaam. 

 Alison van Eenennaam

In some ways cattle work are quite analogous in terms of fertility clinic work what people do for embryo transfer into a female. Versus mice where you’re ding different things.

We just go and get slaughter house ovaries, that’s the source of our oocytes. We don’t care who the mother is, we just want viable eggs.

Versus in a human fertility clinic obviously it’s that lady’s eggs or whoever is trying to get pregnant. There’s a fundamental difference. And if we produce 400 and throw away 390 I don’t really care. Whereas every is precious in a human IVF lab .

 But in terms of the equipment we use, we order from the same companies. We have to sign a thing and say we are not doing human work: they want you to verify that this is not being used for human applications.


With Cosmo the gene-edited bull, the scientists did a lot of sequencing of samples to look through the genome of the zygote. Cosmo has no ‘wild type’ cells, he is not genetically mosaic. Alison van Eenennaam explains what his gene-edited genome reveals.

Alison van Eenennaam [36.44]

No, he has no wild type cells. We did whole genome sequence and PacBio runs. 80x coverage x three, so that’s really high density coverage. There were 11 reads of wild type, and my bioinformatician said that’s just fluff in the air , not legit. He’s biallelic, both chromosomes 17 were cut. He’s what is called a compound heterozygote, the mutations are different on the two alleles, the two homologs. One is a 24-basepair insertion,  an indel, it’s a typical NHEJ repair, like the “ OMG my chromosome was cut, I need to stick it back together.’ That’s what’s typically called an indel, in this case an ‘in.’ The other one is the one that got the SRY-GFP, but it got it in multiple copies and it also got a copy of the backbone. So there’s that.

There’s 80 million reads of the 24 base pair insertion and 80 million of the big guy, which I think ended up being 36 kilobases long in which various orientations. Two of them came together forward forward reverse reverse, it’s an inverted repeat and it looks like there was a kind of recombination after the original insertion. Some of the daughter cells from the original insertion have a smaller insertion that is missing 4 copies. The inverted repeat looks like it looped out. Those cells, a subset, maybe 5% of that subset got that, and all of the other junctions are exactly the same. That’s why I think it’s the daughter of a recombination rather than a separate editing event because the sequences are perfect everywhere except where this thing looped out.

 We did all this preliminary work, we’ve never seen plasmid insert, we look and we’ve never seen it before. And of course the minute we make an animal we get this concatenation and the plasmid insert. That’s fine. For our purposes it’s not a big drama.

It will be used as a big drama probably because of the plasmid insert that was in the polled animals, it would not be acceptable going commercial with an animal. But for research purposes, we got our gene in and that is what we needed to do. It’s not an issue.

I do get a little bit worried about 7 copies because sometimes if you have multiple copies you can get shutdown of the gene. Cosmo is eight months old now. You’ll be interested to know his scrotal circumference this morning was 29.5 centimeters and yes I did the measurement.


Does he let you do that?

Alison van Eenennaam [40:55]

He’s in a cattle shunt so he’s kind of constrained. 32 is what is typically considered mature size scrotum he’s probably a couple of months away from producing semen. If all works well, part of the embryos that we inseminate with him will be green. That will be our first look to know if the gene is still expressing. There is some shutdown of viral promoters sometimes. And he’s got an SV 40 promoter, there’s still some risks. It’s never over till the fat lady sings.

It could be his offspring that inherit that could be xx are just happy girls. It just takes a long time to do these experiments, that’s what research is , it’s waiting for these things to be able to test them.  


So once Cosmo produces semen, she can bank the semen and use it to fertilize eggs in the lab. To see if the gene is expressed in the embryo, the green fluorescent protein will light up. Since Cosmo has multiple copies of the SRY gene, it could be that the gene might no longer be expressed in which case there will be no green fluorescence.

 Perhaps you don’t care about farm animals, you don’t eat beef. That’s ok. But perhaps you care about conservation biology because to help a species come back from the brink of extinction for example, conservation biologists have to find ways to support their reproductive physiology.

Alison van Eenennaam [42:50]

Absolutely, the Revive and Restore group, we have a project with them. They did the Przewalski’s horse that they resurrected from 27-year old cell line. Brought back genetics, went through a nasty bottleneck. They’re working on that. They’re trying to do the white rhino, to clone and gestate that. Knowing what I know to work on domestic species where we know how to super-ovulate them, everything and I look at the rhino people and say good luck. That’s a big ask. Even with all the knowledge we have of reproductive techniques, that’s a big ask.


There are lessons from cattle to be had about gene-editing, about human physiology and conservation biology. Alison van Eenennaam was invited to an experimental biology meeting, which is quite unlike the livestock meeting she otherwise attends.

Alison van Eenennaam

We are totally different communities. There’s a lot of different objectives and aims. There’s anti-animal ag, don’t eat meat people. But we actually do similar work, we just work in different species. It was interesting to be at that meeting to hear about work underway, butterflies. And there are some unusual allies. Groups that have historically, philosophically different, see a shared use of these technologies for restoration and conservation purposes.

Historically, a cattle geneticist/biologist wouldn’t have been invited to a Journal of Experimental Biology meeting. It’s opening up new conversations with Revive and Restore, those types of groups. Some unlikely allies. If you get a bunch of greenies upset about not being able to use this technology. Imagine if they put all of that political power to good instead of stopping things.


Indeed, greenies do get in the way sometimes. Yes, I know, I understand, I don’t necessarily agree with Dr. van Eenennaam here, some environmentalists have important points to make. But of course it’s ok for her to think differently.

I asked her how Cosmo was doing and if his behavior was any different than the behavior she typically sees in cattle.

Alison van Eenennaam  [45.50]

He’s just in our regular herd. he is in a pen with an xx male that was identified through genomic selection. The AI company that he was being tested for sent his genome in to get tested. It came back XX. They thought: ‘oh stupid lab they mixed up our sample,’ sent it in again, came back xx They went and looked, there’s a penis there, what’s going on here. We karyotyped him. He is SRY negative XX male. 


It’s a deletion?

Alison van Eenennaam

SRY is usually on the Y chromosome. A female doesn’t usually have SRY. He’s a male, he’s got a penis, but he’s XX. He is kind of interesting guy, a bit of a control for our females that we get from Cosmo. Because they’ll be SRY positive XX and he’s SRY negative XX. He’s in the pen with Cosmo. He’s missing SRY and Cosmo has 7 copies.


It’s interesting, the transgender community, they talk about continuity.

Alison van Eenennaam [47:20]

I’m always careful to use the term sex, because we’re talking about animals. I never use the term gender. This is a touchy topic. This individual. I’m also doing the similar measurements on him and he’s quite distinctly different from Cosmo. We’re doing hormone assays.

He’s a naturally occurring animal, we didn’t , we found him through colleagues in the AI industry. We have another one coming, too. I think they’re more common than people. So normally a male that is xx and infertile, you chop its head off and eat it. It’s no use for agriculture. People might say, I don’t know why he’s infertile, but off he goes. Maybe that happens 10,000 animals or 100k it might be more common that we know, since people don’t normally karyotype bulls. Since we’ve been working on this project, we’ve identified two. More common. And to your point of fluidity of gender it’s true in all species, these types of things all the time. 


So male and female animals, and yes, I don’t want to say transgender animals but anyway, male and female animals can have a range, a spectrum of, let’s call it identities. Cosmo is XX and has been gene-edited to have an SRY gene on the X chromosome and he has 7 copies of SRY. In his pen is an XX bull who has no SRY gene.

 Gene-editing in cattle is for example relevant to the development of tropical livestock such as through the work of the CTLGH, the Center for Tropical Health Livestock Genetics and Health. That’s an international collaboration focused on ways to improve how livestock can be kept and used for agriculture in the tropics.

Alison van Eenennaam  [49:50]

Even with cattle issues, because of the not the Cartagena protocol, the Nagoya protocol. It makes it impossible to work collaboratively, it’s a double edged sword. I understand the incentive for it was purse. Application is problematic for collaborative work. I advise the CTLGH Center for tropical health livestock. Out of Roslin Institute, funded by Bill and Melinda Gates.

They are working on developing resilient cattle in Africa. And the person, Karen who works with ILRI, the international livestock  institute in Kenya, I think she spends half her life dealing with Nagoya protocol to get access bulls, so we can do much like the 1000 bull genomes, but the hurdles that they have to go through. It’s mindblowing. Are we helping these countries, by doing this or are we making it more difficult for researchers to collaborate. There’s  a balance there that’s not quite being met.


Alison van Eenennaam edits the genomes of cattle and knows of colleagues for example who edit the genomes of horses.

Alison van Eenennaam

Horse people are an interesting crowd.  They have a lot of rules and regulations. If you breed a mare with a thoroughbred, the stallion has to be in sight of the mare for it to be registered. You can’t collect semen and send it to another state and inseminate the mare. That doesn’t count, it’s a way to keep the price of services high. If you can do what the dairy industry does, you get the best bull semen in a semen straw, I don’t need to own a bull. I can get the best semen already. Tried to keep it elite, and supply and demand you know how that works.

There’s a little bit of cloning that goes on, but that tends to be the polo crowd not the racehorse crowd. The polo people seem to believe, cutting horses, and I don’t know if that behavior is particularly heritable they are in the minds of these horse owners. And most horse owners are not agriculturists, they’ve got their money from somewhere else. Money is not object. Whereas in agriculture, money is an object. A gentleman in Brazil is doing myostatin knockout. Everybody does myostatin knockouts. All who do this are male. I don’t know why . that’s just my observation. I don’t know if that is going to make horse run faster. I don’t know if very muscular people being fast runners, I think that is kind of what they are trying to do.

Let’s just say I wanted to edit someone to be a fast runner, human or horse. What gene would I edit, I have no idea. I don’t think that’s a single gene trait. Myostatin certainly is, that’s a gene that makes you muscular. That’s just a great big muscle-y person or horse.

Pharlap, a famous Australian horse, when he necropsied after he died, his heart turned out to be twice as big as a normal horse. That’s what made him run fast. Not big muscles. It’s not obvious to me what you would do to performance enhances a horse from a genome-editing perspective.


Another animal in which some labs explore gene-editing are chickens. Labs are editing the germ cells in chickens.

Alison van Eenennaam [55:00]

Yep, fully reduced to practice and excitingly, combined with surrogate sire, knocking out the germ line. You can basically have a chicken that has no germ cells of its own but has genome edited cells populating the testes in the case of a male and the ovaries in the female.

When those two mate, you get a homozygous animal of a different breed. You can have White Leghorns that have the gonads of a black chicken and Australorp. And when they mate they will produce all Australorp offspring. They will be genome edited for whatever you did. Roslin Institute has done that, Mike McGrew is the researcher there. That’s part of the CTLGH center for Tropical Livestock Genetics and Health


When working with chickens, editing genomes the way she describes using surrogate sire technology, you give an animal gene-edited germ cells. The animal has no eggs or sperm of its own. What gene-editing labs see as an advantage is that there will not be mosaic animals, no animals with differing genotypes in their cells.

Alison van Eenennaam

The beauty of that is that it’s going to be non mosaic. You’re putting in germ cells that have been edited. So there is no chance of mosaicism with surrogate sire technology, which makes it very appealing. There’s certainly a lot of people working on that in mammals. John Oatley had a paper in PNAS earlier this year doing it in pigs and goats. It’s reduced to practice mice. Well they have a picture of one sperm developing in a gonad.  I haven’t seen offspring that are produced from large animals. But they have done this in mice.


Oh, in mice.

Alison van Eenennaam

You can do everything in mice, that’s my hypothesis, it just seems to be more malleable or maybe there’s just more experiments getting done. Everything done in a mouse almost never translates easily across.


Exploratory germline gene-editing is underway in animals such as cattle, chickens and horses. Other animals too.

With people, germline gene editing is not permissible in most countries. Labs are exploring gene-editing in gametes, in sperm and eggs in many animals. And some are thinking about exploring this  approach in people. 

 As of right now, gene-edited animals in agriculture are not allowed to enter the food chain. There is a lot going on in this field and I  plan to keep following it.

And I wanted to add that since doing this interview with Alison van Eenennaam, Cosmo has been euthanized.

The team has collected semen from him. They have used his semen to produce blastocysts and they see that the inner cell mass fluoresces green – or at least 50% of them do, that is in the ones that inherited the GFP on chromosome 17.

As of yet, there are no Cosmo calves, Alison van Eenennaam has not secured the funding for that yet. That would be an experiment in which eggs are fertilized with his sperm and the blastocysts are transferred to surrogate cow to carry a pregnancy.

I spoke with Alison van Eenennaam mainly about cattle but we did also talk about the experiments by He Jiankui whose lab’s work led to three children who were gene-edited before their birth. Badly done human experiments.

Alison van Eenennaam

I am interested as someone who edits embryos to see what he did.


The manuscripts of the work by He Jiankui have not been published in a journal. They are also not on any preprint server, which is where papers can be found that have not been peer-reviewed.

According to sources, the manuscripts were rejected. It is certainly complicated to decide what to do with unethical experiments and records of them. I have a bit more about that in my article and in a forthcoming podcast.

I asked Alison van Eeenennaam about this idea of companies offering parents as a service to help them have babies, to have children with genomes edited before birth. And I played her comments on this earlier in this podcast. It seems to me after hearing a bit about her work, you might hear what she says differently. Here you go:

Alison van Eenennam

I’ve got two pretty good kids, I’m pretty happy with them , pretty happy with what I got the natural way.


And you wouldn’t if you were approached with a brochure gene-editing.

Alison van Eenennaam

I would laugh because the whole notion is ridiculous. The only reason ever would be a homozygous lethal condition that there’s no other way around. But even then. We’re not at that level of precision. I’ve lost a baby I have a still born and I wouldn’t wish that on anybody.

 Let alone 10 recipient women, the whole emotional toll of that. It’s one thing when it’s a surrogate cow, but when you’re talking about a baby and having a loss because editing didn’t go correctly, that’s emotional toll right there.

Maybe a better approach is to fix the gametes. It’s unknown in the embryo but if you can fix the gametes. Look at surrogate sire technologies that are being created in the chicken for example. You can analyze what you’ve done. Make sure there’s nothing there you didn’t want, it’s kind of like cloning. You can check before you produce the individual that you’ve go the edit you want.


That was Conversations with scientists. Today’s guest was Dr. Alison van Eenennaam of the University of California, Davis. And I just wanted to say because there’s confusion about these things sometimes, the University of California Davis didn’t pay to be in this podcast. This is independent journalism produced by me in my living room. I’m Vivien Marx, thanks for listening.

(Credit: Getty Images / Maksymowicz)


Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in