Gene Therapy in Ophthalmology Update 19: A New Virus Vector for Safer Delivery of Gene Therapies

Researchers at the University of California at Berkeley, along with some assistance from the Flaum Eye Institute and Center for Visual Science at the University of Rochester, have come up with a new version of an adeno-associated virus (AAV) vector that can deliver genes deep into the retina using an intravitreal injection of the vector into the vitreous, a less-invasive technique, instead of an intraretinal injection below the surface of the retina, which is the usual way gene therapy is currently delivered.

The study was authored by postdoctoral fellows Deniz Dalkaral (then with Helen Wills Neuroscience Institute of UCal Berkeley, but now with Institut de la Vision in Paris) and Leah C. Byrne (Helen Wills), and graduate students Ryan R. Klimczak and Meike Visel (UCal Berkeley’s Dept. of Molecular and Cell Biology), and Lu Yin and William H. Merigan (Flaum Eye Institute and Center for Visual Science at the Univ. of Rochester), under the direction of Professors John G. Flannery, and David V. Schaffer of UCal Berkeley. The paper, “In Vivo–Directed Evolution of a New Adeno-Associated Virus for Therapeutic Outer Retinal Gene Delivery from the Vitreous”, was published online on June 12th in Science Tranlational Medicine.

As explained by Dr. Jean Bennett, a professor of ophthalmology at the University of Pennsylvania in Philadelphia, who was not involved in the study, but who has done extensive work with gene therapy in the treatment of Leber’s congenital amaurosis, “It shows the results of a very clever system to evolve AAV to target cells in the retina efficiently from an intravitreal injection.”

“Intravitreal injection, whereby a needle is pushed into the eye’s vitreous, or gel-like core, is a common drug delivery procedure performed under local anesthetic in a doctor’s office”, explained Bennett. “But using this routine injection technique in trials of gene therapy for retinal degeneration has thus far proven impossible.”

The problem, as explained by Dr. David Schaffer, a professor of chemical and biomolecular engineering, bioengineering, and neuroscience at the University of California, Berkeley, who led the research, is that current AAV vectors are incapable of penetrating deep into the retina where the target cells for retinal diseases are located. “AAV is a respiratory virus and so it evolved to infect lung epithelial cells,” explained Schaffer, “It never evolved to penetrate deep into tissue.”

Patients receiving gene therapy have theretofore undergone a vitrectomy (removal of the vitreous) and a direct intraretinal injection, which requires hospitalization and general anesthetic, and can sometimes even damage the retina. “If it were possible to inject AAV into the vitreous instead of the retina and still get gene delivery to the target cells, said Bennett, “one could envision the [doctor saying], ‘Ok, well just come into the office and get your gene therapy, tomorrow afternoon at two.’”

With that aim, Schaffer and colleagues used a process called “directed evolution” to randomly create millions of variations of the AAV virus to determine which ones were better at tissue penetration. They injected regular AAV into the vitreous of mouse eyes and one week later collected photoreceptor cells from deep within the retina. The tiny percentage of AAV vectors that made it into those cells were then amplified, repackaged into virus particles and injected into the vitreous again. They repeated the injection, recovery, and amplification a total of six times, finally isolating 48 AAV variants for sequencing. Two thirds of those isolates turned out to be the same variant, and Schaffer and colleagues named it 7m8.

Lastly, to determine whether the 7m8 vector would be likely to show similar deep penetration in the human retina, Schaffer had the vector carrying a gene-encoding fluorescent protein injected into the vitreous of macaque eyes. Primate retinas are considerably thicker than those of mice, and the vector did not consistently reach the deep cell layers – showing a spotty penetration pattern rather than the wide and even pan-retinal penetration that had been seen in the mice. However, 7m8 did effectively target photoreceptor cells of the fovea – a thinner part of the primate retina that is essential for the sharp detailed vision humans use when reading and driving. “That’s a really important region to protect,” said Schaffer. “For the quality of life of patients who are going blind, if you can at least protect the fovea that would be a huge improvement.”

Eye cells labeled with green fluorescent protein have successfully taken up the virus, showing that the ‘evolved’ virus (right) is more effective than the virus currently used for gene therapy (left). The new virus is particularly good at targeting the critical photoreceptors (top layer). (Source: University of California, Berkeley)

Note: All of the “directed evolution” work with mice to discover the 7m8 AAV vector was done at UCal Berkeley, while the confirmatory primate injection and imaging was done at the Univ. of Rochester.

Schaffer and colleagues don’t yet know what makes the 7m8 vector so much better at tissue penetration than its AAV ancestor, but they plan to find out and use that knowledge to further improve its penetration in the primate retina.

"Building upon 14 years of research, we have now created a virus that you just inject into the liquid vitreous humor inside the eye, and it delivers genes to a very difficult-to-reach population of delicate cells in a way that is surgically non-invasive and safe. "It's a 15-minute procedure, and you can likely go home that day."

The engineered virus works far better than current therapies when administered from the vitreous of the eye in rodent models of two human degenerative eye diseases (X-linked retinoschisis and Leber’s), and can penetrate photoreceptor cells in monkeys' eyes, which are similar to those of humans.

Schaffer said he and his team are now collaborating with other investigators to identify the patients most likely to benefit from this gene delivery technique and, after some preclinical development, hope soon to head into clinical trials.

Schaffer predicts that the viruses can be used not only to insert genes that restore function to non-working genes, but can knock out genes or halt processes that are actively killing retinal cells, which may be the case in age-related macular degeneration.

As noted by  Dr. Stephen Rose, Ph.D., chief research officer, Foundation Fighting Blindness, one of the co-funders of the research, “This is a critical next step in the development of retinal gene therapies. The enhanced AAV holds potential for treating more of the retina and doing so more safely. Incremental advancements like this are essential to getting the best treatments out to the patients.”

The investigators showed efficacy for the 7m8 AAV in a large animal as well as mouse models of retinoschisis and Leber congenital amaurosis, or LCA (RPE65 mutations). In the mouse studies, the virus was able to penetrate the retina and deliver a corrective gene to enable the retina to function normally.

While the large animal did not have a retinal disease, the virus transduced many regions of its retina. Ultimately, in both types of animals, the AAV was able to deliver genetic cargo to a variety of retinal cells, including: photoreceptors, the cells that provide vision; the retinal pigment epithelium, a layer of cells providing nutrients and waste disposal; and ganglion cells, which are a target for emerging, vision-restoring optogenetic therapies. (Editor’s Note: See, for example, my writeup of the “Nirenberg Technique”, an optogenetic approach to restore near normal vision to the blind.)

Most notably, the intravitreally administered AAV was able to penetrate the fovea, a small pit in the center of the retina rich in cones, which provides the vision most critical to daily living, but is often made fragile by degenerative diseases. Researchers have been concerned that injections underneath the fovea could cause permanent damage and vision loss in patients with advanced degeneration in their central retina.

AAVs are currently used for gene delivery in several retinal gene therapy clinical trials, including those that have restored vision in children and young adults with LCA (RPE65). AAVs are attractive for gene delivery because of their natural ability to penetrate a variety of cells. In addition, humans are exposed to the virus in nature and, therefore, tolerate it well.

To identify the optimal AAV for intravitreal gene delivery, the scientists used a process called “directed evolution” to randomly create millions of variations of the virus. The variants were then screened in mice to identify the top candidates for gene delivery to the retina. In addition to looking for an AAV that could penetrate retinal cells well, the researchers searched for a variant that could pass through a formidable barrier in the eye known as the inner limiting membrane, or ILM, which separates the vitreous from the retina.

The scientists from UC Berkeley plan to work with others to perform additional toxicology and efficacy studies to ready the 7m8 AAV for study in humans.

Sources:

In Vivo–Directed Evolution of a New Adeno-Associated Virus for Therapeutic Outer Retinal Gene Delivery from the Vitreous, Deniz Dalkara1 et al, Science Translational Medicine, June 12, 2013

Genes Get in Your Eye: Directed evolution of a gene therapy virus vector improves its penetration into the retina, Ruth Williams, The Scientist, June 12, 2013

Easy and Effective Therapy to Restore Sight: Engineered Virus Will Improve Gene Therapy for Blinding Eye Diseases, Robert Sanders, University of California - Berkeley, June 12, 2013

Researchers Identify Better Virus for Retinal Gene Delivery, Foundation Fighting Blindness, June 12, 2013