Modern gene therapy is gradually moving away from a single universal idea of "fixing damaged DNA." Sometimes a working replacement can indeed be delivered to a gene. And sometimes there is no need to change the sequence at all — it is enough to lift the molecular prohibition that prevents the existing gene from working.
Two recent studies illustrate this difference well.
The first is devoted to KCNV2 retinopathy — a rare hereditary retinal disease. The KCNV2 gene encodes the Kv8.2 protein, which participates in the functioning of potassium channels in photoreceptors. In pathogenic KCNV2 variants, the function of rods and cones is disrupted: visual acuity decreases, and photophobia, color vision disturbances, and night blindness appear. There is still no specific treatment for the disease.
The researchers used an adeno-associated viral vector — AAV — to deliver an optimized working copy of human KCNV2 into photoreceptors. This is not editing of the damaged gene: the existing mutation remains in place, while the cell receives an additional instruction that allows it to produce Kv8.2 again.
After subretinal administration of the therapy to mice lacking functional Kv8.2, the electrical responses of the retina improved markedly. At certain doses, the visually guided responses of the animals under both dim and bright lighting approached those of healthy mice. In retinal organoids grown from the cells of a patient with KCNV2 retinopathy, the researchers also restored the production of the Kv8.2 protein.
For now, this does not mean that anyone has managed to "restore vision." What is meant is a preclinical proof of principle: the method works in an animal model and in human cell models. The next question is considerably more difficult — whether it will prove safe and effective enough in patients.
The researchers applied a completely different approach to Prader–Willi syndrome.
Here the problem lies not only in damage to an individual gene. In Prader–Willi syndrome, there is no normal expression of a group of genes from the paternal copy of the 15q11–q13 region of the chromosome. The maternal copy of many of these genes exists, but is silent because of genomic imprinting — an epigenetic system that determines which parental copy of a gene should be active.
It was precisely this silence that the scientists attempted to lift.
A team from Japan used the CRISPR/dCas9-SunTag-TET1 system. Unlike ordinary Cas9, dCas9 does not cut DNA. It is used as a molecular navigator: it is directed to the required region of the genome and recruits the TET1 enzyme, which is capable of altering DNA methylation.
The researchers acted on the PWS-ICR imprinting center in cells obtained from patients with Prader–Willi syndrome. Demethylation made it possible to switch on again several genes of the maternal chromosome, including SNRPN, SNORD116, IPW, and MAGEL2. It is especially interesting that the changes persisted after these cells were converted into hypothalamic organoids — three-dimensional models of tissue involved in the regulation of hunger, hormonal functions, and metabolism. In the organoids, the overall gene expression profile was also partially normalized.
But here an important subtlety came to light.
After the epigenetic "awakening" of the region in neuronal cells, UBE3A-ATS was also activated — an antisense transcript capable of suppressing UBE3A. In the edited organoids, the researchers did indeed see a decrease in UBE3A expression. The authors warn: excessive suppression of this gene could theoretically create changes associated with the pathology of Angelman syndrome.
That is precisely why this work is especially illustrative. The epigenome cannot be perceived as a series of independent switches. Sometimes switching on one region alters the functioning of an entire chain of neighboring genes.
Both strategies are still far from routine clinical therapy. For KCNV2, long-term efficacy and safety in humans must be confirmed. For Prader–Willi syndrome, the task is even more difficult: it is necessary to learn how to deliver the editor to the right brain cells, to control the strength and duration of the effect, and not to disturb the delicate balance of other imprinted genes.
But the direction is already visible. In one case, medicine adds to the cell the genetic instruction it lacks. In the other, it tries to restore the voice of a gene that has remained in the DNA all along but was switched off.
This is precisely what is becoming one of the most interesting features of the new molecular medicine: a disease can be treated not only by rewriting the genome, but also by changing the way the cell reads it.
