Like a spy on a secret mission, the #9 serotype of adeno-associated virus sneaks through the blood-brain barrier, delivering genes to astrocytes and other cells of the central nervous system. Scientists at Nationwide Children’s Hospital in Columbus, Ohio, found that intravascular AAV9 crosses from the bloodstream into the nervous system, and yields widespread transduction of CNS cells. The results are published online today in Nature Biotechnology. While scientists have transduced brain and spinal cord cells before, it generally requires direct injection into the appropriate region, and transgene expression tends to stay tightly localized. The current study is the first to show viral gene transfer across a broad swath of the nervous system.

“It allows us to make genetic manipulations on a wide scale,” said first author Kevin Foust of Nationwide Children’s. Translating the technique to therapy for human patients is one application, of course, but the authors noted AAV9 could also have an important place in basic research. For example, scientists could use the virus to make a quick-and-dirty transgenic mouse that expresses new genes in the nervous system. The mechanism by which AAV9 traverses the blood-brain barrier is still unknown, so the study opens up many basic science questions about its gate-crashing abilities.

The paper is likely to generate some excitement, said Pedro Lowenstein of Cedars-Sinai Medical Center in Los Angeles, California. “If it’s true, people are just going to jump on this and you are going to see lots of papers using this technique in the future,” Lowenstein said. “They can get really massive transduction of the brain by injecting the viruses into the bloodstream. No one has achieved this to that degree before.”

Adeno-associated viruses come in a variety of flavors—more than 50 serotypes, by one estimate (Gao et al., 2004). They make an appealing candidate for gene therapy because they are not known to cause any pathology. AAVs steal in under cover of another infection such as a cold and, scientists think, install their genome in the nucleus, but separate from the host’s chromosomes. Then they “sit really quietly” and can persist for a long time, said principal investigator author Brian Kaspar, also at Nationwide Children’s.

AAV9 has a distinct capsid that must, in some way, allow it to cross over into the nervous system. In preliminary studies, Foust also tried serotypes 6 and 8 for CNS transduction. “It was nothing compared to AAV9,” Kaspar said. They hypothesize that AAV9 is able to traverse the epithelial cells that fence off the nervous system. Then, Kaspar said, the virus would run up against astrocyte projections. Perhaps, the scientists muse, AAV9 can bind to a receptor on those astrocytes and thus gain entry to the inner sanctum.

Foust and colleagues engineered AAV9 carrying a GFP gene under the strong chicken β-actin hybrid promoter commonly used in such experiments. In treated animals, transduced cells lit up sections of the nervous system like a Christmas tree. They injected viruses into the tail vein of both adult and neonatal mice. Day-old pups showed widespread transduction in the brain as well as motor neurons and some glia. The blood-brain barrier was probably not fully solidified in these animals, allowing the virus to leak into the nervous system. In adults, astrocytes primarily expressed the transgene, although the occasional spinal neuron and a patch of hippocampal neurons also produced GFP.

Foust used a high dose of viral particles, ranging from 400 billion to four trillion viruses per mouse. Scaled up for human size, that’s one big batch of virus, noted Matthew During of Ohio State University in Columbus, and might be more than a body could handle. “That viral load is so dramatic that I think the virus would be floating all over the body,” he said. “That amount of virus almost invariably causes immune response.” It could also prove toxic to the liver, he said.

However, the study authors noted that the paper is only the first step toward human therapy. “To translate this to a human is certainly going to take more work,” Kaspar said. In clinical trials, it would probably be possible to target the virus more effectively to the central nervous system. While the tail was the most easily accessible injection site for day-old mice, people are a bit bigger and have many spots where a catheter or needle could fit. For example, doctors could inject the virus into an artery or vein that leads directly to the brain or spinal cord, allowing the viruses a crack at the CNS before being filtered by the liver.

Two prime disease targets for AAV9 therapy would be amyotrophic lateral sclerosis and spinal muscular atrophy (SMA). Many experiments have shown that in ALS, astrocytes play a major role in the degeneration and death of motor neurons (Yamanaka et al., 2008). AAV9 could potentially deliver genes to fix ailing astrocytes. SMA, another fatal motor neuron disease, is the second most common genetic disorder causing childhood death. Again, AAV9 might be able to deliver the missing protein that the cells need, perhaps very early in life. It is not certain exactly when the human blood-brain barrier seals completely, Kaspar said, but “There may be a window [during which] one could diagnose, and still treat.”

Other nervous system disorders are potential candidates for AAV9 therapy. “We like to think that the astrocyte targeting could be utilized in adult diseases to deliver trophic factors or therapeutic factors throughout the entire central nervous system,” Kaspar said. In this case the astrocytes would become part of a delivery system, pumping out appropriate growth factors or anti-inflammatory molecules, for example, that cannot cross the blood-brain barrier on their own. “Find a disease that responds to a trophic factors and it becomes a candidate,” Foust said.

For researchers, the ability to change gene and protein expression in an animal’s nervous system opens up myriad possibilities. Foust noted that it takes a year or two to engineer a transgenic mouse. With AAV9, a scientist could use transduction to make a mosaic out of a wild-type pup, and then make an educated guess about what a fully transgenic animal might be like. “You can have at least a hint of an idea as to whether you need to make that two-year investment,” Foust said.

The time for AAV9-based human therapies has not yet come; clearly, there’s more work ahead. But the blood-brain barrier now has a gate, and scientists have the key. “At least we have a virus that can do the trick,” Lowenstein said, and that’s the first step.—Amber Dance


  1. The results reported by K.D. Foust et al. are of great interest for a number of reasons.

    They provide evidence for a surprising transduction pattern in the central nervous system (CNS) following intravenous administration of recombinant AAV9. In newborn mice, transduction of long-projection neurons following intravenous injection presents an overall pattern similar to what has previously been observed for other AAV serotypes such as AAV6 (1). However, the rate of transduction reaches unprecedented efficacy, with about 60 percent of motor neurons transduced. Even more surprising is the dramatic change in the transduction pattern observed following a similar procedure in adult mice, with a predominant targeting of astrocytes. This is at odds with the tropism of most AAV vectors, which has proved mostly neuronal so far. When injected directly into the brain parenchyma, the recombinant AAV9 used in this study displays the expected neuronal pattern of infection. Thus, rather than reflecting a peculiar tropism of the serotype 9, astrocytic transduction seems to be due to a combination of specific conditions including the intravenous route of administration and the adult stage of the recipient mice.

    Until now, most strategies based on peripheral injection of viral vectors have been based on retrograde transport along axonal processes to enter the CNS. Quite remarkably, recombinant AAV9 appears capable of crossing the blood-brain barrier (BBB), which is usually not permeable to viral particles. Within the CNS, astrocytes may plausibly be the primary cell type exposed to viral particles. Nevertheless, it is important to find out by which mechanism the virus is capable of crossing the BBB, in order to determine whether that particular virus can be safely employed in these conditions, and whether this feature could be conferred to other types of viral vectors.

    Genetic modification of astrocytes via the peripheral injection of a viral vector opens new avenues for manipulation of glial cells, whose role appears increasingly important in a number of neurodegenerative diseases. In amyotrophic lateral sclerosis, such a flexible technique may be crucial to decipher the interplay among astrocytes, neurons, and possibly also microglia leading to the demise of neuromuscular connections (2). However, before considering any therapeutic application, it will be essential to assess the robustness of this transduction pattern across rodent disease models and various methods of vector preparation. For instance, it is unclear whether the use of self-complementary AAV vectors constitutes an important factor (3). Most importantly, it remains unknown how this approach can be scaled up to large mammalian species such as primates, and whether the virus doses needed are compatible with clinical applications.


    . Systemic AAV6 delivery mediating RNA interference against SOD1: neuromuscular transduction does not alter disease progression in fALS mice. Mol Ther. 2008 Jun;16(6):1018-25. Epub 2008 Apr 15 PubMed.

    . ALS: a disease of motor neurons and their nonneuronal neighbors. Neuron. 2006 Oct 5;52(1):39-59. PubMed.

    . Efficient retrograde neuronal transduction utilizing self-complementary AAV1. Mol Ther. 2008 Feb;16(2):296-301. PubMed.

    View all comments by Patrick Aebischer

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Clades of Adeno-associated viruses are widely disseminated in human tissues. J Virol. 2004 Jun;78(12):6381-8. PubMed.
  2. . Astrocytes as determinants of disease progression in inherited amyotrophic lateral sclerosis. Nat Neurosci. 2008 Mar;11(3):251-3. PubMed.

Further Reading


  1. . Recombinant adeno-associated viral vectors as therapeutic agents to treat neurological disorders. Mol Ther. 2006 Mar;13(3):463-83. PubMed.
  2. . Recombinant AAV-mediated expression of human BDNF protects neurons against cell apoptosis in Abeta-induced neuronal damage model. J Huazhong Univ Sci Technolog Med Sci. 2007 Jun;27(3):233-6. PubMed.
  3. . Hybrid baculovirus-adeno-associated virus vectors for prolonged transgene expression in human neural cells. J Neurovirol. 2008 Nov;14(6):563-8. PubMed.
  4. . Delivery of AAV-IGF-1 to the CNS extends survival in ALS mice through modification of aberrant glial cell activity. Mol Ther. 2008 Jun;16(6):1056-64. PubMed.
  5. . Adeno-associated viral (AAV) serotype 5 vector mediated gene delivery of endothelin-converting enzyme reduces Abeta deposits in APP + PS1 transgenic mice. Mol Ther. 2008 Sep;16(9):1580-6. PubMed.

Primary Papers

  1. . Intravascular AAV9 preferentially targets neonatal neurons and adult astrocytes. Nat Biotechnol. 2009 Jan;27(1):59-65. PubMed.