An Update on the Use of Gene Therapies to Convert Retinal Cells

A few years back, researchers reported on a novel approach to treating the degenerative blindness of retinitis pigmentosa, an inherited condition in which the rod photoreceptor cells responsible for low-light and peripheral vision become progressively more dysfunctional, eventually leading to the death of other retinal cells. The researchers found that tinkering with levels of Nrl in retinal tissues can make rod cells transform into something more like the cone cells responsible for color vision. In normal development of retinal photoreceptor cells, those with a lot of Nrl become rods, while those with less become cones, but as demonstrated by this team, adult photoreceptors can be coerced into taking on the character or one or the other. While this isn't as good as fixing the underlying problem that causes rod cell dysfunction, it so far appears to be a potentially beneficial approach.

This is far from the only situation in the human body in which some form of conversion of adult cells might be helpful as a palliative or compensatory treatment, in absence of a true cure that addresses root causes. There are many examples of specialist cell populations impacted by aging or disease, arising from a common progenitor and thus closely related to surrounding cells. Dopaminergenic neurons, islet cells, and so on through a long list of cell types. I don't think that researchers should be starting out with this sort of approach as the end goal of their work medicine - that end goal should be to fix the underlying cause of the problem and thus effect a complete cure. But if it becomes feasible along the way, then why not?

Today I noticed a recent update on this line of research, now moved to CRISPR-based gene therapy in mice. Given the advent of CRISPR and the great reduction in the cost and difficulty of gene therapies, I have to wonder why there isn't more of a thrust towards treating the genetic cause of this condition. While the number of specific mutations thought to relate to the condition is quite large and varied - it isn't a nice, neat single gene inherited disease - it is no longer enormously costly and challenging to deliver correct versions of multiple genes to just the retina. In many ways retinal conditions are as close to an ideal testing ground for gene therapies as you are likely to find in the human body, given the relative isolation of the eye from other tissues, and the distinctive differences in those tissues that enable accurate targeting via a number of methods. Still, people more familiar than I with the costs have decided that a more general and compensatory approach makes sense.

CRISPR-Based Therapy Prevents Retinal Degeneration

Retinitis pigmentosa, which affects around one in 4,000 people, causes retinal degeneration that eventually leads to blindness. The inherited disorder has been mapped to more than 60 genes (and more than 3,000 mutations), presenting a challenge for researchers working toward a gene therapy. The results of this latest study suggest that a broader, gene-editing-based therapeutic approach could be used to target many of the genetic defects underlying retinitis pigmentosa. "This combination of CRISPR technology with an adeno-associated virus vector, a system tried and true for delivering genetic information to the retina, may represent the first step in a global treatment approach for rod-mediated degenerative disease."

Researchers designed a CRISPR single guide RNA (sgRNA) to target a retinal transcription factor, Neural retina leucine zipper (Nrl), which specifies rod cell fate during retinal development and maintains rod cells within the mature retina. The team delivered the Nrl-targeted sgRNA and the Cas9 endonuclease directly to the retina of mice on two separate adeno-associated virus (AAV) vectors. The advantage of an AAV vector is the ability to maintain long-term gene expression in non-dividing cells, such as those of the retina.

Retinitis pigmentosa causes gradual cell death - first of rod cells, responsible for night vision, followed by the more scarce cone cells, which enable color and daylight vision. Rod cells also provide structural and nutritional support to cone cells. Researchers showed that targeting Nrl could preserve the functions of cone cells in mice. Using a Cre-based recombination system, the team found that eliminating Nrl could partially convert rod cells into cone-like cells, preventing their deaths as well as the secondary cone cell death seen in retinitis pigmentosa. "The idea is that we do not turn rods into real cone cells, but for the rod cells to gain some cone feature so that they will resist the mutation effects that cause them to eventually die."

Because the Cre-recombination approach is not readily translatable into humans, researchers looked to the CRISPR system instead. The team first injected the retinas of 2-week-old wild-type mice with the experimental therapy vectors, then observed the mice for three to four months. The researchers saw reduced Nrl expression in the animals that received the therapy compared to mice injected with control vectors. At six weeks after injection, the treated animals' rod cells had downregulated rod genes and upregulated cone genes. The treatment did not result in any deleterious effect on cone cells, the researchers reported.

Nrl knockdown by AAV-delivered CRISPR/Cas9 prevents retinal degeneration in mice

In retinitis pigmentosa, loss of cone photoreceptors leads to blindness, and preservation of cone function is a major therapeutic goal. However, cone loss is thought to occur as a secondary event resulting from degeneration of rod photoreceptors. Here we report a genome editing approach in which adeno-associated virus (AAV)-mediated CRISPR/Cas9 delivery to postmitotic photoreceptors is used to target the Nrl gene, encoding for Neural retina-specific leucine zipper protein, a rod fate determinant during photoreceptor development.

Following Nrl disruption, rods gain partial features of cones and present with improved survival in the presence of mutations in rod-specific genes, consequently preventing secondary cone degeneration. In three different mouse models of retinal degeneration, the treatment substantially improves rod survival and preserves cone function. Our data suggest that CRISPR/Cas9-mediated NRL disruption in rods may be a promising treatment option for patients with retinitis pigmentosa.


""This type of 'imprecision medicine approach' could be the first application in any patients with retinal pigmentosa due one of the 67 genes," Tsang told The Scientist."

Rather than painstakingly working out the interactions between various mutant and non mutant versions of all these genes, could they just use a couple (or one) CRISPR treatment to change all 67 genes to non mutant versions in one go?

George Church's lab edited 62 retroviruses out of the pig genome at once late last year smashing the previous record of 6 edits at once:

Posted by: Jim at March 14th, 2017 8:55 PM

Maybe it's simply a business strategy. Maybe patenting/marketing/approving a different therapy for each of the 3,000 mutations combinations is too costly/cumbersome and they prefer a one-size-fits-all approach than a personalized medicine approach, even if the outcome is inferior.

Posted by: Antonio at March 15th, 2017 5:56 AM

I've since remembered that adding a new gene in after a CRISPR edit is still a low efficiency process. George Church's group just slashed and disabled 62 genes. If someone can figure out how to get CRISPR to insert new genetic material efficiently in the body, that really would be a revolution (low efficiency methods work in cells in culture dishes because you can select the small number of successfully transformed cells).

Posted by: Jim at March 15th, 2017 10:45 AM

Post a comment; thoughtful, considered opinions are valued. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.