Remember these letters

Alexander Ershov, N+1 

Recently it became known that the UK will be allowed to edit the genome of human embryos for research purposes. For this purpose, the CRISPR/Cas9 technology, which was discovered just a few years ago, will be used. We tried to answer the most obvious questions that arise in this regard: what is it, why is it necessary and how will the new technology change medicine. 

Until now, such studies in the United Kingdom and in the West have generally been banned. Earlier, about a year ago, the first experiments were conducted in China, but their legal status was unclear and they caused a flood of criticism from researchers. The UK will be the first Western country to officially allow the use of genome editing technology in relation to human embryos.

It is worth noting that the permission applies only to research purposes. It has been issued so far to the only scientific team – a group headed by Kathy Niakan from the Francis Crick Institute. Scientists will be required to destroy the resulting GM embryos within 14 days of receiving them. And, of course, they cannot be planted for a woman to bear. 

Potentially, CRISPR/Cas9 technology can change the attitude of mankind to hundreds and thousands of hereditary diseases. If earlier they were either completely incurable, or allowed palliative, symptomatic treatment, now it is possible to treat them "for real", that is, to eliminate the very cause of the disease. 

Simultaneously with the advent of genome editing technology, the possibility of its "improvement" appears, in a variety of senses. So far, we are talking about fairly simple (from the point of view of the inheritance mechanism) diseases, but potentially not only "broken" genes can become targets for editing, but also genes simply associated with an increased risk to health. Or even genes responsible for harmless physiological features like the ability to drink milk in adulthood or success in sports.

What is commonly called "cancer" is a giant family of various diseases with different mechanisms of occurrence. There are types of cancer, the probability of which is closely related to particularly "unsuccessful" variants of some genes. A typical example is the BRCA1 gene, mutations in which can increase the likelihood of breast cancer several times. Potentially, using CRISPR/Cas9 technology, it is possible to make changes to the genome of a sperm or egg and thus prevent the transmission of a mutant variant of the gene to their children.

The problem is that for most oncological diseases, heredity does not play a big role, which means that genome editing technology will be almost useless. On the other hand, there are severe hereditary diseases that have high heritability, but it is so complex and confusing that it is not clear where and what changes should be made to the genome to reduce the risk of their occurrence. A typical example is schizophrenia, the risk of which is believed to be inherited by 80 percent (this is shown in identical twins). At the same time, the molecular mechanism of inheritance of schizophrenia was completely incomprehensible until very recently and has only now begun to become clearer.

If we talk about the fact that using CRISPR/Cas9 it will be possible to treat in the first place, then these are primarily simple monogenic diseases like beta-thalassemia, cystic fibrosis or hemophilia.

Now more and more bioengineers are switching to this technology. However, there is one fundamental difference between old and new technologies: this is the direction of making changes. It is in this that the fundamental difference between CRISPR/Cas9 technology lies.

Previously, in order to achieve the appearance of a new desired property in the body, bioengineers simply embedded a DNA construct into cells. At the same time, it was impossible to predict the place in the genome where this construct would fall (except in some cases, like baker's yeast). This led to the fact that, firstly, the natural version of the gene in the genome was preserved (if it was there, of course) and only supplemented with a new, artificial version.

This method is suitable for obtaining some new property, for example, enhanced production of growth hormone in GM salmon or for the synthesis of vitamin A in rice grains. However, when it comes to replacing a broken gene with its correct copy, especially in human DNA, it is clear that non–directionality is a big minus. In addition, accidental insertion into the genome can lead to inefficient operation of the transgene – the activity of any gene in nuclear organisms depends on its environment, on the local structure of chromatin. Therefore, a transgen trapped in an unsuccessful piece of the genome may simply be turned off or, conversely, too active. Unlike the old methods, CRISPR/Cas9 technology allows not just to embed a new sequence in DNA, but to replace its old version with a new one.

First, a special nuclease (i.e., an enzyme that cuts DNA) introduces a double-stranded break in the right place of the genome. The nuclease finds this place with the help of a short guide RNA (selected by scientists), whose sequence must match the desired sequence in the genome with the letter. After the rupture is made, the internal mechanisms of the cell, the so-called repair system, are activated.

It should be understood that the appearance of a double–stranded break in DNA is an emergency situation for any cell. The gap leads to mutations and generally threatens the integrity of the genome. Therefore, there are special proteins that find "broken ends" in the genome and trigger a "repair" reaction. The gap, of course, can simply be glued back together, but this is fraught with the loss of several "letters" at the junction and, as a result, a shift in the reading frame and a complete shutdown of the gene. Therefore, the cell usually prefers to find a similar sequence nearby in the genome and use it as a sample to restore the correct sequence at the site of the break. This is where the enzymes can be given the DNA variant that we want to replace the natural sequence with.

The difficulty with editing the genome so far has been precisely to make this gap. It should appear in a single place of the genome and nowhere else - precisely because such breaks lead to the appearance of mutations. For comparison, the size of the human genome is about three billion nucleotides, and the guide sequence of RNA, which should find its landing place in the genome, has a length of about twenty to forty nucleotides. It's amazing that she can do it at all. If we are not talking about a single cell, but about gene therapy of an entire tissue, then the task becomes even more difficult – all cells must be modified, but each only once.

Prior to the discovery of the CRISPR/Cas9 system, scientists had already tried to develop methods for making directional breaks in DNA. For example, our former compatriot Fyodor Urnov has done a lot of work in this direction. We are talking about the rational design of nuclease proteins that would independently (without a guide RNA) find unique sequences in the genome. The difficulty with these methods is that they require the development of their own protein for each specific task, which then needs to be synthesized, isolated, tested, etc. It is much easier to work with a universal nuclease and a specific guide RNA, but scientists did not know about this possibility until the bacterial immunity system was discovered.

It lies in the fact that a huge number of bacteria carry in their genome (where, it would seem, everything has been clear for a long time) an elegant system of adaptive immunity against viruses. The basis of this system is special sections of the genome – short palindromic cluster repeats or CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats). 

Repeats act as "shelves" between which "dossiers" on viruses that the ancestors of this bacterium once encountered are located in the genome. "Dossiers" are simply short fragments of DNA that match in sequence with fragments of the DNA genome of viruses. If a virus with matching DNA gets into a bacterial cell, it will be recognized fairly quickly by a special enzyme, the Cas9 nuclease. The latter uses an RNA copy synthesized with CRISPR to search for viral DNA. 

If any fragment of the virus genome coincides exactly with what is recorded in the "dossier", Cas9 cuts the viral DNA and starts a chain of reactions, as a result of which it is all destroyed. In general terms, this scheme resembles RNA interference, which was discovered in nuclear organisms ten years earlier, but this (like everything in eukaryotes) is a significantly more complex and less efficient system.

At the beginning of this year, encouraging data appeared on the treatment of Duchenne myodystrophy in adult mice, and experiments were conducted in three different laboratories independently. Just a few days ago it became known about the successful application of technology for the treatment of severe retinitis pigmentosa.

The startup Editas Medicine, closely associated with the pioneers of the technology, has already attracted more than $ 120 million in investments (including from Google). This money will be used to create an experimental treatment for Leber's amaurosis of the tenth type – this is hereditary blindness associated with damage to one of the genes necessary for the operation of light-sensitive retinal cells. Clinical (i.e. human) trials in Editas Medicine promise to begin next year.

This sounds like a meaningless alarmism, usually coming from the mouth of opponents of GMOs, but in fact the situation here is fundamentally different.

The efficiency of editing with CRISPR/Cas9 is not yet sufficient to talk about "scalpel–precise" genome correction - no matter what the authors of popular publications write. Simultaneously with the necessary gap in the genome, superfluous ones are often introduced, and this, as already mentioned, provokes mutations. Even if the break is made correctly, the efficiency of homologous recombination, due to which the original sequence is replaced by the desired one, is very far from 100 percent.

What is the real effectiveness is a more complicated question than it seems, because it strongly depends on the type and nature of the cells in which the editing is carried out. What works well in mice may not work well in humans. And until researchers start working with real human embryos and eggs, one can only guess about the effectiveness of the procedure and the level of accidental ruptures.

To date, there are results of only one experiment with genome editing in a human embryo – the same ones that were published by a Chinese group in April last year (and rejected by Science and Nature on ethical grounds). Then scientists worked with 86 fertilized eggs, of which 71 survived and 54 were selected for analysis. In 28 out of 54 cells, the Cas9 enzyme introduced the necessary breaks into the genome, but only in four cases the repair of the gap ended with the replacement of the gene sequence with the desired one. At the same time, scientists discovered multiple breaks in the genome of cells where they should not be.

Such low efficiency and a high level of errors turned out to be a surprise for the authors of the work themselves, as they honestly admit in the article. Whether this low efficiency is due to the "crooked" hands of scientists or to the peculiarities of human embryos will be unclear until the experiments are repeated many times by other groups. Until today, when the UK finally allowed them to be conducted, Western researchers did not have such an opportunity.

Much has been done in this direction since the publication of the Chinese work. For example, in December last year, scientists managed to create an artificial version of the Cas9 enzyme, which is many times more accurate than the natural one and almost does not introduce unnecessary breaks into the genome.

It will be more difficult to increase the efficiency of sequence replacement, since it relies entirely on natural mechanisms of homologous recombination, but work is underway in this direction. However, even if the effectiveness remains low, in the absence of side effects, CRISPR/Cas9 technology can still be used to make inherited changes in the human germ line. For example, you can take connective tissue cells from a patient, edit the genome and select only those of them where the editing took place without complications. These cells can be used to produce induced stem cells, from which spermatozoa can then be obtained and used in IVF. There are difficulties here, but at least this technology works on animals.

But not everything is so rosy on the CRISPR horizon. The closer the actual clinical application of the technology is, the more heated the dispute about who will receive income from it. According to some estimates, the cost of an exclusive patent for a technology can reach many hundreds of millions of dollars (at least in such amounts, the volume of venture financing of CRISPR/Cas9 startups is measured). The patent dispute over CRISPR/Cas9 promises to be louder than anything that has ever happened in the field of intellectual property in biotechnology.

On January 11 of this year, the US Patent and Trademark Office (USPTO) began the procedure for checking patents related to CRISPR/Cas9 for "interference". Officials will have to determine which of the research groups holding similar patents should be given priority in creating the technology: publications, witness statements, postal correspondence and records in laboratory journals will be used. The future of the entire technology will depend on the outcome of the process, because the rightful owners will be able to simply prohibit the use of their technology by competing companies, and this, in the end, will put an end to the hopes of rapid implementation of CRISPR/Cas9 in the clinic.

Scientists who initially jointly tried to bring the technology to mind were divided into at least two opposition camps, each of which claims to be the priority of discovery. On the one hand, this is Jennifer Doudna, who together with Emmanuelle Charpetier published a key work on the practical application of Cas9 in genome modification. This article was published at the end of 2012. In the spring of the following year, Dudna filed a patent for this technology, but in the same year, many similar papers appeared from other researchers who tried to improve the method in their own way. One of them, Feng Zhang from the Broad Institute, filed his own patent for CRISPR/Cas9 in October of the same 2013. And although this happened after the filing of Dudna's patent, Chzan's patent passed through a simplified procedure and was issued first.

Now a major artillery has been used in the patent dispute: Eric Lender, an MIT professor and one of the co-chairs of the Committee on Science and Technology under the US president, recently published an article in Cell "CRISPR Heroes", in which he sets out his view on who made the greatest contribution in this whole story and why. What caused Lender's impulse to sort out this issue right now – the desire to influence the patent office or purely academic interest – is unclear. It is quite expected, however, that he (as the founder of the Broad Institute, from which Czan filed his patent) does not attach as much importance to the contribution of Dudna and Charpentier as the latter would like. It is clear that Dudna and Sharpentier, no matter how big the academic and hardware weight of the Lender, will not give up without a fight. Just look at their comments on the ill-fated article, which they have already left in Pubmed. They can be understood, because it's not only and not so much about the ill-fated patent. Of course, we are talking about who will get the next Nobel Prize.

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

03.02.2015