Researchers Have Just Found A Better Way To Edit DNA

Illustration for article titled Researchers Have Just Found A Better Way To Edit DNA

Sorry CRISPR, but there's a new genomic editor in town — and this one's better than you. It's a new approach to site-specific gene targeting that will allow scientists to safely replace disease-causing genes with functional copies. And they've already used it to relieve the effects of hemophilia in mice.


One of the most important biotechnological breakthroughs of the last several years is a genomic editing approach called CRISPR/Cas9. It allows researchers to replace a faulty, or mutated, version of a gene with a working copy. The technique is relatively easy to use, it can be used to modify multiple genes, and it features quick turn-around times. Currently, scientists are using it to create specific strains of lab animals, but in future it could be used for gene drives and to treat genetic diseases in humans. More controversially, it could conceivably be used for human enhancement and to confer entirely new characteristics altogether.

CRISPR, however, is not without its drawbacks. It requires the co-delivery of an enzyme called an endonuclease to snip the recipient's DNA at specific locations, and it relies on the co-insertion of genetic "on" switches called promoters to activate the new gene's expression. These lead to an increased chance of adverse effects, including cancer and various DNA abnormalities.

Now, Stanford researcher Mark Kay has found a new way to go about genome editing that avoids these problems. The new technique is considered safer and longer-lasting than other methods. Kay and his colleagues demonstrated the new technique by enabling mice with hemophilia to produce a missing blood clotting factor. From the Stanford release:

The technique devised by the researchers uses neither nucleases to cut the DNA nor a promoter to drive expression of the clotting factor gene. Instead, the researchers hitch the expression of the new gene to that of a highly expressed gene in the liver called albumin. The albumin gene makes the albumin protein, which is the most abundant protein in blood. It helps to regulate blood volume and to allow molecules that don't easily dissolve in water to be transported in the blood.

The researchers used a modified version of a virus commonly used in gene therapy called adeno-associated virus, or AAV. In the modified version, called a viral vector, all viral genes are removed and only the therapeutic genes remain. They also relied on a biological phenomenon known as homologous recombination to insert the clotting factor gene near the albumin gene. By using a special DNA linker between the genes, the researchers were able to ensure that the clotting factor protein was made hand-in-hand with the highly expressed albumin protein.


The integration of the clotting factor gene was key to the successful experiment; other clinical trials involving hemophilia have relied on the expression of a free floating, unintegrated gene in the nucleus.

The real issue with AAV is that it's unclear how long gene expression will last when the gene is not integrated into the genome. Infants and children, who would benefit most from treatment, are still growing, and an unintegrated gene could lose its effectiveness because it's not copied from cell to cell. Furthermore, it's not possible to re-administer the treatment because patients develop an immune response to AAV. But with integration we could get lifelong expression without fear of cancers or other DNA damage.


Amazing, we're inching closer to the day when this technology will finally be made available to people suffering from virtually any kind of genetic disorder.

Read the entire study at Nature: "Promoterless gene targeting without nucleases ameliorates haemophilia B in mice".


Image: ktsdesign/shutterstock



I'm not convinced that this that exciting. It certainly can't replace CRISPR in a large number of situations. Just off the top of my head, the corrected disease is special because

a) It involves a factor in the blood. Albumin is a very highly expressed gene in the liver, it is true, and it has extensive interactions with blood, which means that the missing clotting factor can be produced in high quantities in the liver and then released along with the Albumin into the bloodstream where it goes to work and spreads to all the blood vessels in the body. What if we are targeting something that is going wrong with, say... the lungs? Cystic Fibrosis is a common first-round candidate for genetic manipulation, but the problem there lies in the epithelial cells in the lungs. Those cells do not express much albumin, so you'd have to tie it to some other gene, but there aren't really any massively expressed lung-specific genes, to my knowledge, so you'd get some pretty massive potential off-target effects.

b) It is a 'simple' recessive disorder and can be fixed by a transgene. The method described here is not really as... elegant... as CRISPR. In CRISPR, you have the option to physically replace a defective copy of a gene with a working copy. Here, you are just adding an extra copy, and that solves the problem. But other disorders are far more complicated. Oftentimes the mutant copy acts as a poison pill and resists the transgene by still incorporating into complexes and still causing the abberant behavior. This method is unable to address those sorts of disorders.

c) The method they are using, directed homologous recombination, is DREADFULLY inefficient. Just look at their 'successful result' here. They targeted their gene to albumin. Albumin has a concentration in the blood of ~40 mg/ml. F9 weighs about 10% less than Albumin, so if we had perfect recombination in every cell in the liver producing albumin, we should get ~36 mg/ml of F9. Instead, we get (being very generous) 300 ng/ml of F9. That means that for every ONE cell that integrated the F9, there were roughly 120,000 cells which did not. Again, by tying their gene to the most highly expressed protein in the blood, they were able to restore 20% of the expression of a gene which is normally expressed over a 100x less than Albumin. For this disease, where even that 20% is enough to ameliorate the problem, everything is fine, but what about some other disorder in which you need the majority of your cells to have made the change? CF won't get better if 1/120,000th of your epithelial cells are now working normally.

All in all, its cool, and definitely has a number of very important applications (and deserves its spot in nature), but a CRISPR-killer, it isn't.