Genetic editing legacy – update on the first designer babies
Designer babies: birthed in controversy
The creator of first live-birth designer babies was sentenced to three years in prison yesterday!
It has been just over a year since the news of the world’s first genetically designed babies was announced in China. Apparently this announcement was not done by a “rogue scientist”, but rather a young, very well-connected and aspiring scientist who was not only interested in the accepted genetic editing methods but also seemingly hiding a secret - that he had produced the first genetically engineered embryos used to give birth to live children! It shocked the world and then condemnations swiftly followed.
But an historic event it was! And Dr. He Jiankui is now, for better or for worse, recorded in history forever.
After the avalanche of criticism it seemed everyone even remotely associated with the project or the researcher was trying to distance themselves as much as possible and thus there never was an official, sanctioned scientific publication of this momentous genetic experiment on human beings.
Now in December 2019, the same scientific news outlet that first broke the story of this event, has published a leaked copy of the designer babies manuscript that was apparently to be submitted for scientific publication. The revelations are indeed stunning, and not in a positive light at all. We thought it would be good to look into this and stitch this story to another genetic editing episode from the recent past that was much cooler but barely covered in the popular press.
Assuming this manuscript is real of course, there are immediately two things egregiously wrong with this experiment that should have been dealt with before any allowed human experimentation. It is not like we are doing some mere genetic investigative test. We are talking about purposeful editing of our genetic code in human beings which can then be passed on to future generations. This was carried out without full understanding of what the consequences may be!
But before we get into the two horrible problems with the work done (or rather “not done”), let’s recap what was done. Number of different human eggs and sperm were collected from multiple couples. The egg and sperm of a specific couple were used for an in vitro fertilization - so the insemination of an egg inside a test tube instead of a womb. Then this artificially inseminated egg wa s allowed to grow into a tiny embryo - a small ball of cells really - and the whole thing was then exposed to the CRISPR genetic editing technology. CRISPR/Cas9 allows for DNA modification at a desired location. Dr. He and his team decided to focus on the CCR5 gene known to be involved in promoting HIV infection . A known mutation in humans who are resistant to HIV infection, have a specific segment of DNA deleted from the CCR5 gene. This mutation, known as CCR5Δ32, is what was attempted to be created in these first designer babies.
The first and biggest problem with the overall experiment is that it did not confirm if the CRISPR/Cas9 technology was accurate in its genetic editing outcome. This would require doing detailed experiments with other early stage embryos, where every cell had its DNA decoded. CRISPR/Cas9, while turning the genetic world upside down and being the most-attention grabbing DNA editing system we can think of, might not penetrate every cell to accomplish its mission. Plus in those cells that it will affect it can also cause edits in the DNA in places that are unintended and this is called off-targeting. This is the primary knock against its effective use in clinical practice currently and for all its promise, it is still very unreliable.
What the authors of the designer baby paper did, was take a cell or two from the embryo and investigate it. Apparently everything was awesome! But we cannot know if ghastly mistakes might have happened in any other targeted cell. Maybe it was just a lucky draw from the deck of cards that a good cell was selected. What should have happened is a number of early embryos should have been investigated cell by cell, if you were trying to be the first person ever to apply the use of this approach on a future human being that will have to bear all the consequences (and potentially even their children). As far we see it, it seems like high chances of some editing going astray. We just do not know what off-target alterations might have taken place and now there are going to be real life consequences for the genetically designed twins.
Early mutation equals big impact
If you establish purposeful genetic mutations early in the development, they can result in profound effects. For all intents and purposes, when introducing a mutation that early in the game of development, you might as well say “you were born with it” - as if you got it from one of your parents. We get mutations in our DNA all the time, in every cell, and some do not get fixed. We are a symphony of mutated DNA all over our bodies as a consequence of the constant cellular insults to our body over time. Mutations just need time to eventually pop up. By the time they do, they may have a diminished effect, depending how often the cells with a mutation end up being replicated. In such situation, the mutation spread is localized (think of your moles for example, or other skin changes that are permanently present). Because mutations just need time to even happen, that is why, as we age, you are going to see so many more of those skin changes. And just like your skin, the same thing is happening everywhere in your body! That is simply part of aging. Accumulating DNA mutations.
But first we go through a process of development and growth and so the timing of mutations can have profound impact as any DNA mutation before our development and growth will be more widespread in future body parts than DNA changes that happen way past our complete maturation. Once development is over, the DNA mutations can be dangerous, but they will be localized to a tiny environment out of which they simply will not have much opportunity to spread to many daughter cells. If that area of the body is not growing much, that DNA mutation has not much of a chance to be passed around.
Thus you can see how important those early DNA mutations in the development can be, because they can be propagated to entire swaths of your body, entire organ systems, and thus result in very serious conditions and diseases. This is not anything out of the ordinary. In fact, there is a term for it: mosaicism – when you present with more than one genome type in your body. We can get some of the basic estimates from the non-invasive prenatal testing (NIPT) outcomes, which focus on DNA mutations in a developing fetus. For example, the egg fertilized by sperm (that is pretty much all that a sperm cell is doing, just providing some informational genetic fertilizer) will end up producing both the fetus and the placenta. In the NIPT test, you do not actually assess the fetal DNA, but rather the placental DNA. In the NIPT test, you test mother’s blood which does not come in contact with fetal DNA. It comes in contact with placental DNA (as placenta cells die and release their DNA into the mother’s bloodstream).Thus placental DNA should be identical to fetal DNA. But it is not always. In rare occasions, placental DNA will be different from that of the fetus. It means that a mutation in DNA occurred early in the development to produce genetically different fetus from the placenta!
Thanks to detailed NIPT test analysis, we have an idea how often that happens. If the placenta carries a DNA mutation that is not in the fetus, that will contribute to NIPT test false positive result. Taking the more conservative estimate, NIPT false positive result rate is about 0.1%. That would mean that at least 1 in 1000 live births will have a DNA mutation in placenta that will not be present in the actual baby upon birth. Those are live births! Imagine what the rate might actually be if we were able to capture the DNA mutation rate behind all miscarriages and stillbirths! Yeah, probably a lot higher! It is believed that around half of all miscarriages occur due to gross chromosomal alterations. We do not even know to what degree individual DNA point mutations might be a contributing issue but we can be certain there are some as we observe point mutations in diseases more frequently than structural mutations.
This would suggest that random mutations in early development can potentially carry deadly consequences, and we should therefore be damn sure before we are introducing any mutations in the embryo of a human being that will be born one day, as well as that we do NOT introduce a mutation that might be detrimental to them. Like, say - anything causing premature death as just one example.
This type of confirmation was never carried out!
Technically this could never be carried out on the genetically edited embryo that is being implanted in the uterus. Not with current technology because in order to check the DNA of an early embryo cell, we need to destroy the cell to get to the DNA. So we can check one or two cells perhaps and no more. The rest of the embryo cells will be unchecked for any DNA alterations, as they are needed for future development.
The next best approach would be to show in other embryos that off-targeting does not take place so that we could confidently assume it would also not take place in the embryo that is implanted in a woman.
By investigating this issue thoroughly, we would hope to be able to at least say that based on past studies, we can say with 98% confidence that no additional mutations are introduced because in the experimental embryos, and we have seen such off-target mutations only 2% of the time. Or whatever these numbers would be for ethically acceptable levels to proceed.
The reason why we see this as so ethically wrong, is that even if the expected DNA alteration went exactly as planned, what other parts of DNA might have been affected is just unknown, and we will not know until we can collect more data on the designer babies as they grow. Because of how poor the current state of CRISPR/Cas9 system is at the moment, these children are condemned to never feeling certain if their future health outcomes are a consequence of some side DNA mutations or not.
CRISPR/Cas9 weakness: not all it is cracked up to be
Don’t think this is likely? Think again! Not long ago one paper shed light on how badly inaccurate CRISPR/Cas9 system might be, and not just due to “off-targeting” but even at the site of where the mutation is supposed to happen! Here the authors only introduced the Cas9 enzyme with guide RNA to elicit DNA cutting at desired site. Template DNA was not provided (for CRISPR/Cas9 background see our prior post). This alone was enough to frequently introduce mutagenesis on that very spot where DNA was cut, most frequently deletions of DNA material, some even very large! Plus other unusual rearrangements were taking place. If the Cas9 with its guide RNA was targeted to a specific gene, the loss of that gene product due to introduced mutations was very high, from 60% to nearly 100% of the time! What is even weirder, is that 17% of the time, they observed DNA rearrangements away from where the Cas9 and guide RNA was targeted.
Thus unwanted mutagenesis is more common than you would first think based on the fact that a decision to proceed with experiments on human beings was made. But the amount of unwanted mutagenesis should be a fact affecting the decision to do these experiments! Well, we do not have much of a sample size with regards to humans. These were the first two CRISPR babies ever produced. But we have an incredible amount of data for similar attempts in mice. One recent paper showed the difficulty of genetic editing. Granted, they looked at a more complex approach where two separate, identical mutations were attempted to be introduced into mice DNA, but the numbers are still quite revealing. The data was collected from 20 different facilities around the world, totalling generation of 17,887 CRISPR-exposed zygotes (fertilized eggs), so that is big data set to look at. Of these, 12,764 were transferred into recipient mouse females, resulting in birth of 1718 pups. This in turn provided access to 1684 live mice that were subsequently analyzed for alterations to their DNA. This already should give you a hint of the difficulty of the process.
Here we go. Only 15 mice had the correct desired DNA alteration, or 0.87%! 40% had DNA editing events! Of these 13% had insertion of one of the desired mutations, 11% showed DNA being deleted between the two desired sites of where the mutations were supposed to go, and 16% showed any other type of DNA editing that also had nothing to do with desired outcome. This was assessed across 56 different genes that were being targeted.
So based on this information, is one to think that nothing bad might have happened to the DNA of those kids? Looks less and less likely!
And there is evidence to suggest that is exactly what happened!
One of the persons who happened to review the manuscript was Dr. Kiran Musunuru. He noted that in the DNA sequencing data from the placenta and umbilical cord of designer babies after their birth, there was evidence of overlapping DNA information. This means that in the decoded sequence, in the background in same position as some DNA nucleotide, there was also presence of another. This might have been dismissed as background noise, but it could also mean that simply other mutations were introduced elsewhere, and the babies are a combination of different mutations just in that CCR5 gene alone! Meaning, they are “mosaic”, more than one type of genome now makes up their body.
Dr. Musunuru saw the experiment as unethical from the moment the designer babies were announced.
The second biggest issue with the work performed on designer babies was the fact that the mutations that were introduced in these babies were in fact NOT identical to the ones that are found in the normal human population. The attempted CCR5Δ32 deletion mutation was not successfully duplicated at all, and yet the team elected to proceed with implantation. In other words, DNA mutations for which we have no clue what the consequences will be were introduced into these designer babies. But considering that CCR5Δ32 mutation is the one observed in that HIV resistant phenotype, this suggests that there is actually very little room to play when it comes to genetic tinkering of the CCR5 gene, and again, time will tell if what has been altered will have any potential negative effects.
Unfortunately the MIT Reviews article did not disclose what these mutations were, so we still do not know what was really done to the designer babies in their CCR5 genes. We just know it ain’t a CCR5Δ32 deletion! Further confirming the point made above that CRISPR/Cas9 is not all that it is cracked up to be! It makes mistakes!
Genetic editing, where are we headed?
And what has happened to Dr. He? Nobody knew until yesterday! Yesterday it was announced he was sentenced by a Chinese court to three years in prison, a large fine, and a life-long ban from working with reproductive technology for his alleged crime of illegally practicing medicine. He was found guilty of forging documents and concealing the true nature of his work to both the patients as well as doctors involved. Prior to this, it was total secrecy on the topic from China. He has been missing for a year and there has been no additional information about the first designer babies, or the outcome of other pregnancy that was already underway when the announcement of birth of the girls was made.
That is just some crazy stuff! Hiding the nature of such experiments from those involved in them is as bad of a crime as it gets in a scientific world.
But, it is also hard to believe that one person could pull that off (ok, two others involved in the scheme were also sentenced at the same time).
That is because it should be super hard to be even possible. There is a stringent oversight built into the scientific processes, especially for those involving human subjects. You would have to fool a huge number of people along the way, and forge tons of information. Almost too much to be able to get away with it, especially in a country like China which is now world famous for its advanced surveillance capabilities.
What will be the long term consequences of the unethical development of these first designer babies?
The correct approach of studying the use of the technology will certainly continue, although some might worry it will be slowed down. Somatic genetic editing will continue, at least to attempt to treat people already born with conditions, in an effort to minimize the disease symptoms. This is definitely already taking place and moving forward. Apart from the DNA editing human cases we already discussed in a past post, the technology has recently been demonstrated for the first time in the US, with CRISPR/Cas 9 used on cancer patients.
We are certain this technology will be used again to make more designer babies once we know how to use it safely in the eradication of genetic diseases. Some are really clamoring for it, like Dr. Denis Rebrikov who is experimenting on on editing out DNA mutations responsible for certain hearing loss conditions.
The potential is indeed enormous.
One example is the concept of being able to repair specific tissues. We wanted to present one such example that captured our attention a few years back as it is also a good way to provide some further background behind the complexity of genetic editing and what is involved. This article showcased effective gene modification therapy targeted to a specific tissue in adult mice.
Of course, once again, the CRISPR/Cas9 genome editing tool was the star. It does the tough work at a DNA level, but another big problem is how do you get it there? How do you bring the system to the DNA inside the cells of a tissue so that it can perform the desire edits?
Taxi to nuclear DNA
There are two competing carrier systems being developed: viral carriers and lipid nanoparticle carriers. These scientists chose to use both, but not how you would expect! These experiments were sure not for wimps and why we wanted to discuss it.
First lipid nanoparticles, were a cumulative achievement of the trial and error of many scientists (like everything else that is described here). All of our cells are surrounded by a tiny film of a lipid layer (or in more plain language, layer of fat), called the cell membrane. Thus lipid nanoparticles mimic what is found in biology but on a vastly simpler and smaller scale with a touch of novelty to them. In the first of its kind, these spherical lipid carriers were used to deliver mRNA Cas9 to liver cells of mice. By delivering mRNA, which is short lived, it ensured that existence of Cas9 protein was going to be only temporary, reducing the threat of undesired genome cutting. Pretty clever!
Then the viral carriers. The truth is gene therapy via genome editing has actually been around for quite some time in clinical trials. The first documented attempt was in 1990, and the first reported successful demonstration of genetic editing was in the year 2000, using viruses. Since then a number of successful examples exist of gene therapy treating a specific disease, for example: anemia, severe combined immune deficiency, or genetic blindness. Currently there are over 1500 different gene therapy or gene transfer clinical trials around the world according to https://clinicaltrials.gov website, including an entire gamut of human afflictions, from cancer to skeletal and cardiac muscle tissues to a variety of neuronal-based conditions. Safety concerns with the use of such complex procedures has still precluded gene therapy as standard clinical care, but you can see that the interest is immense.
Adeno-associated viruses are an example of such virus carriers. They are tiny, non-pathogenic DNA viruses that consists of only two genes. They are almost too simple, and require other helper viruses just to propagate. Their viral genetic material can be substituted with any DNA of therapeutic interest, and this has been one of the standard methods used for clinical gene therapy. It was this system that was used to deliver the guide RNA for Cas9 and a repair DNA template.
The final outcome was delivery of a system in mice engineered to mimic human hereditary tyrosinemia that leads to severe liver damage. Treated mice showed that the mutation responsible for the disease was reversed in some of the liver cells, also reversing disease symptoms such as weight loss and liver damage. Truly a remarkable achievement and an important breakthrough for the future of therapeutic genome editing.
This type of science should be talked about as it is bound to stir up many different opinions, especially if it arrives suddenly in use without any understanding of where it comes from. If the public has a poor understanding of what is actually being achieved, a life saving procedure could face opposition due to miseducation.
For now, we are still nowhere near treating people with genome modifications as a routine procedure, with understanding of the genome function still being the priority, and luckily, both fields are rapidly advancing. Merogenomics aims to bring access of genome sequencing technology to interested individuals, but if the full promise of genomics is to be achieved, public education and awareness about this burgeoning field of science is also necessary.
This article has been produced by Merogenomics Inc. and edited by Jason Chouinard, B.Sc. Reproduction and reuse of any portion of this content requires Merogenomics Inc. permission and source acknowledgment. It is your responsibility to obtain additional permissions from the third party owners that might be cited by Merogenomics Inc. Merogenomics Inc. disclaims any responsibility for any use you make of content owned by third parties without their permission.
Products and Services Promoted by Merogenomics Inc.
Select target group for DNA testing
Or select popular DNA test
Pharmaco-genetic gene panel
Non-invasive prenatal screening
Cancer predisposition gene panel