Measure it seven times

What could be wrong with edited human embryos

Polina Loseva, "The Attic"

In early August, an international team of scientists reported on the successful editing of a human embryo: they managed to cut out a mutant copy of the gene responsible for hereditary heart disease. However, less than a month later, refutations began to appear: another group of scientists suggests being more careful about the results obtained and requires more detailed evidence. We are looking into what could have gone wrong.

The experiment on editing embryos did not arise from scratch: it was preceded by a long working out of the technique on mouse and human embryos. However, it was a group of scientists led by Shukhrat Mitalipov who managed to achieve the desired result: most of the embryos obtained did not contain mutations in any of the cells. Here is their recipe for "fixing a person":

1. We take a mature egg with a normal copy of the gene and a sperm with a mutant copy of the gene.

2. We inject the sperm nucleus into the egg (as in in vitro fertilization). We also inject Cas9, an enzyme capable of cutting DNA, into it. And finally, we add a probe – a section of RNA that is complementary (that is, capable of sticking) to the mutant version of the gene.

3. We're waiting. At this time, the probe sticks to the mutant gene in the sperm nucleus, and Cas9 cuts it at the ends from both DNA strands. In one of the chromosomes of the sperm, a through hole is obtained.

4. We are waiting for three more days. During this time, the embryo manages to divide several times, now it has 4-8 cells.

5. We take each cell separately, extract DNA from it and check which version of the gene is contained there.

In the Mitalipov group experiment, 72% of the embryos turned out to be completely healthy (carrying only a normal copy of the gene), and the remaining 28% were heterozygotes (carrying both copies of the gene).

We cannot know for sure what exactly happened in the fertilized embryos: the transformation of DNA is not visible in the microscope, we can only guess. Therefore, there may be differences in the explanation of what happened. According to the authors of the study, since mutant sequences were not found in most embryos, therefore, Cas9 cut them out and the cells got rid of them. As for the holes formed in the chromosomes, two options are possible: either the edges of the holes are randomly stitched (but part of the adjacent genetic material is often lost) – the so-called non-homologous connection of the ends, or the whole chromosome (in this case, the maternal one) comes to the rescue – homology repair is taking place. That is, the maternal chromosome is separated into two strands in the area where the gene under study is located, and information from the gene is copied to the paternal chromosome. As a result, both chromosomes carry a healthy copy of the gene, and if this is the case, then genome editing can be considered successful. The authors believe that this second mechanism took place in the embryos.

Types of repair of double-strand breaks in DNA. Blue and black are homologous chromosomes. In the areas of rupture, the repair proteins are marked with colors. Image: Hannes Lans et all., 2012.

But a group of scientists from the USA and the UK have a different point of view on what happened, which they write about on the Biorxiv website (although this site is not reviewed by scientists, so the materials presented on it theoretically may contain unreliable information, but in this case the authors make quite logical comments on the original study):

Homology repair is quite rare in humans. Until recently, it was generally believed that this mechanism did not work in human cells until it was discovered in tumor cells. Mitalipov and co-authors themselves note in their article that the stem cells on which they tested the experimental technique practically did not use one chromosome to repair the other.

As far as is now known, the maternal and paternal nuclei (or rather, the pronuclei, the precursors of nuclei) are strongly spaced in the space of the zygote. They do not merge and do not form a common core, but directly proceed to the first division. At first, they double, still being at different poles, and converge only later, when the chromosomes line up in the center of the cell, forming a spindle of division. Therefore, it is unclear when the maternal chromosome could be so close to the paternal one to "help" her patch up the hole.

Thus, critical scientists question the conclusion of Mitalipov's group. If repair due to homology is unlikely, then the chances are growing that a non-homologous connection of the ends has occurred and the mutant gene is simply lost. And this can no longer be considered editing the embryo.

The answer to some of these comments is found in the original article itself. The authors of the study believe that the germ cells and the zygote have unique DNA repair mechanisms that are not typical for other cells of the body. They attribute this to the fact that DNA in germ cells is of particular value and it is necessary to repair its breakdowns without loss of material. At the same time, they say nothing about the problem of the convergence of the paternal and maternal pronuclei.

There is another way to find out how it really was – to prove directly that the chromosome break was repaired in one way or another. To do this, the authors of the original article analyzed DNA sequences located near the chromosomal hole. If the cells used a non-homologous connection of the ends to repair the rupture, then neighboring sequences would most likely suffer. However, no damaged sequences were found in the edited embryos. From this, the authors concluded that it is precisely the use of the maternal chromosome.

However, in the refuting article, this conclusion is doubtful. Scientists-critics write that with a non-homologous connection of the ends, the damage could affect longer sections of chromosomes than those that were tracked during the experiment. And if the paternal chromosome loses such a site, then it will remain in the cell only in one copy – in the maternal one. And with further analysis, it turns out that everything is in order, because the maternal site is in place, and the paternal one is simply not there.

"We recognize that these results should be confirmed by additional studies, and we urge other scientists to reproduce our findings," Shukhrat Mitalipov said in a letter to The Attic. Nevertheless, he continues to stick to his position. "Our findings and conclusions are based on an elaborate experiment conducted on hundreds of human embryos. The criticism ... does not contain any new results, but, on the contrary, relies on an alternative explanation of our results based on pure speculation," the scientist said. Now Mitalipov is preparing a detailed scientific answer on these issues.

The key issues raised during this discussion are fundamental, despite the apparent complexity. It's not just about the molecular mechanisms that work in this particular case, but about a possible breakthrough in the field of reproductive technologies. If the Mitalipov group's technique proves successful and safe, it will be possible to rid children of hereditary diseases with its help. However, if their results are really contradictory, then such manipulations seem extremely risky. For example, if a large section of the paternal chromosome is lost with a non-homologous connection of the ends, then the embryo then carries only the maternal copy of this section. But what if this maternal copy is defective? The consequences for the body can be unpredictable. The risk is great, but the stakes are also high. We will look forward to developments on the new front of reproductive technologies.

Portal "Eternal youth" http://vechnayamolodost.ru 12.09.2017