Improving adeno-associated viruses for gene therapy delivery

February 12, 2021

New strategy to cloak AAVs from unwanted immune responses

Two microscopy images of retinas.
Retinas of pigs that received an unmodified (left) or engineered AAV vector (right). The modified AAV prevented immune cells (white) from infiltrating into the photoreceptor layer (green). Credit: Sean Wang/HMS Genetics Department and Blavatnik Institute


One of the best ways to package gene therapies and deliver them into cells is to use adeno-associated viruses (AAVs). This delivery vector does not cause disease and can efficiently target many different cell and tissue types. Indeed, the U.S. Food and Drug Administration recently approved AAV-based gene therapies to treat spinal muscular atrophy and a form of inherited retinal dystrophy. However, AAVs can sometimes lead to unwanted immune responses, so Harvard Stem Cell Institute (HSCI) scientists developed a way to modify AAVs and lower their risk.

The study was a collaboration between HSCI Affiliate Faculty member George Church at Harvard’s Wyss Institute for Biologically Inspired Engineering and Harvard Medical School (HMS), and HSCI Principal Faculty member Constance Cepko at HMS. Because the AAV genome can activate an immune response via the protein Toll-like receptor 9 (TLR9), the researchers developed a “coupled immunomodulation” strategy where TLR9-inhibitory sequences were incorporated directly into the AAV genome along with therapeutic DNA sequences. When tested in different tissues in mice, as well as eye tissues in pigs and non-human primates, the approach showed broad anti-immunogenic potential. The results are published in the journal Science Translational Medicine.

Cloaking AAVs using coupled immunomodulation

“We hypothesized that small snippets of DNA that bind and inhibit TLR9 activation, including DNA sequences from the ends of human chromosomes called telomeres, would be a way to cloak the AAV genome from this immune-surveillance mechanism when incorporated directly into it,” said Ying Kai Chan, who is the first- and co-corresponding author of the study and a former postdoctoral fellow in the Church lab.

The team generated a series of synthetic “inflammation-inhibiting oligonucleotide” (IO) sequences that inhibit TLR9. When directly incorporated into the AAV genome, the IOs decreased immune responses in human immune cells, compared to an unmodified vector. Next, the researchers administered AAVs as a systemic treatment or locally into muscle tissue in mice. Control viruses lacking IO sequences induced an immune response both in the livers and in muscle tissue. These immune effects were blocked or much reduced in mice that received engineered AAVs containing IO sequences. Moreover, the gene delivered by the engineered AAVs showed enhanced expression, indicating potentially higher efficacy.

Investigating coupled immunomodulation in the eye

The eye is often described as an immune-privileged site because of the presence of a blood-retina barrier that limits entry of immune cells, as well as immune-suppressive factors. However, multiple clinical trials have reported intraocular inflammation following the delivery of therapeutically relevant doses of AAV into the eye, demonstrating a limit for immune privilege. Although most AAV-based gene therapies in the eye are directly applied to the retina (subretinal injection), AAV delivery to the vitreous cavity (intravitreal injection) of the eye would be less invasive and potentially allows therapies to target more cells — but it is highly inflammatory.

The researchers applied their AAV modification strategy to intravitreal injections in mice. Using in vivo imaging and immune cell characterization techniques, the team found that incorporating IO sequences reduced the inflammation and numbers of infiltrating T cell populations in the eye compared to unmodified AAVs. This also coincided with a boost in the retina’s expression of the delivered gene.

Next, the team tested this strategy in large animals, first in pigs via subretinal injections and then in monkeys via intravitreal injections. In pigs, unmodified AAVs resulted in the shortening of photoreceptor cells and a corresponding loss of high-acuity vision. The modified AAVs prevented this situation, as well as reversed the infiltration of the retina’s photoreceptor layer by immune cells. The modified AAVs’ immunosuppressive effects were not as pronounced in the monkeys that received intravitreal injections, although the coupled immunomodulation approach delayed inflammation and allowed a two-fold increase in the expression of a therapeutic gene.

“The results from the intravitreal toxicity induced by AAV, and the modest response to the TLR9 blocking sequence and to steroids, indicate that there is more than one mechanism leading to toxicity from this injection site. We can now go forward with this understanding and search for additional pathways,” said Cepko, the Bullard Professor of Genetics and Neuroscience in the Blavatnik Institute at HMS and an Investigator of the Howard Hughes Medical Institute.

“Every novel therapeutic modality that achieves initial success in the clinic has to grapple with emerging issues before it can be deployed broadly, and AAV gene therapy is no exception. Our work represents a critical step in development of next generation AAV vectors that are safer and more effective,” said Chan.

Read more

The original version of this story was published on the Wyss Institute’s website on February 10, 2021, under the title, “A new vision for AAV-delivered gene therapies.”

Source article: Chan, Y. K. et al. (2021). Engineering adeno-associated viral vectors to evade innate immune and inflammatory responses. Science Translational Medicine. DOI: 10.1126/scitranslmed.abd3438

The study was funded by the Wyss Institute for Biologically Inspired Engineering at Harvard University, National Institutes of Health, European Research Council, National Eye Research Centre, The Underwood Trust, and Ally Therapeutics.