Primarily, it is believed by many of us that the use of Adeno-associated virus (AAV) as a vector for gene delivery will be more favorable than other vectors. Secondly, the serotypes of AAV may be utilized to fill many niches of gene transfer where one serotype can only fill a few. Finally, the standardization of some aspects of serotype production described in this manuscript allows for a more direct evaluation of the in vitro and in vivo performance of distinct serotypes.

  Does it describe a new discovery of a new methodology that's useful to others?

The manuscript describes a methodology for the production and purification of AAV serotypes. It demonstrates how to utilize knowledge gleaned over two decades of experimentation with AAV serotype 2 (the most studied serotype) so that the other serotypes can be tested under similar conditions more efficiently.

  Could you summarize the significance of your paper in layman's terms?

AAV2 is a nonpathogenic virus that has been developed as a gene transfer vector, recombinant AAV2 (rAAV2). The only viral DNA carried by rAAV2 are two 145-base terminal repeats (TR) that are required for replication and packaging of the foreign gene. Most transgenes have been developed with AAV2 TRs, but not all serotypes efficiently package transgenes with AAV2 TRs. Part of the significance of this paper is the development of AAV serotypes that efficiently package AAV2 TR containing transgenes. This was accomplished by recombining the replication gene of the serotypes to contain some domains from AAV2,

for replication, and another domain from the other serotype for packaging. In this way the same transgenes in different serotypes can be directly compared with the only difference being the shell of the virus. Additionally, these serotypes can now be used as platforms for experiments designed at understanding which domains on the shell are responsible for differences between them.

  How did you become involved in this research?

As a postdoc in Jude Samulski's lab I have been interested in manipulating the shell of AAV serotypes to determine what domains of the shell are responsible for the differences seen in in vivo expression. Using the serotypes in their original design required building new transgene vectors to package serotype specific vectors. T.R. Jude suggested that building new serotype vectors would be more efficient. From the literature I knew that some of the serotypes would require more manipulation to package AAV2 TRs. Smith and Kotin (1998) and King et al., (2001) showed that the small replication protein was required to move the DNA into the virion. While Yoon et al., (2001) showed that the first 232 amino acids of the AAV2 replication protein were all that was required for binding AAV2 TR. The information from these papers pointed the way for the sucessful development of these serotype vectors that allow for the efficient packaging of AAV2 TRs.

Joseph E. Rabinowitz, Ph.D.
UNC Gene Therapy Center
The University of North Carolina at Chapel Hill
Chapel Hill, NC, USA