1. ¹INSERM UMR 649, CHU Hôtel Dieu, Nantes, France
2. ²Etablissement Français du Sang Pays de la Loire, Nantes, France
3. ³Department of Molecular Genetics and Microbiology, University of Florida, College of Medicine, Gainesville, Florida, USA
4. ⁴Center of Excellence for Regenerative Health Biotechnology, University of Florida, Gainesville, Florida, USA
5. ⁵Department of Pediatrics, University of Florida, College of Medicine, Gainesville, Florida, USA

Correspondence: Philippe Moullier, moullier@gtc.ufl.edu; Richard O. Snyder, rsnyder@cerhb.ufl.edu

To the Editor:

Recombinant adeno-associated viral (rAAV) vectors are increasingly being used in the clinic,¹ and more clinical applications for rAAV-mediated gene transfer are being developed in preclinical settings. To share and compare pharmacodynamic, pharmacokinetic, toxicologic, and efficacy data from nonclinical and clinical studies performed by laboratories using different vector–transgene combinations, it would be necessary to determine—across these studies—vector strength, potency, and particle-to-infectivity ratio in terms of equivalent titer units to obtain a common dosage unit. There is a lack of standardization and inability to compare titer values between preclinical and clinical studies for vectors made in different laboratories. Therefore, a highly characterized reference standard stock (RSS) of rAAV is required to facilitate these comparisons. An RSS would allow researchers to normalize their titer values to the common standard, thus allowing each group to state its titers in units that could be compared to those used in other studies; this comparison could also happen retrospectively.

The AAV reference standard efforts were initiated to address the lack of normalization of doses administered to animals and humans.² Today, an AAV serotype 2 (AAV2) and an AAV serotype 8 (AAV8) RSS are being produced; they will be characterized by several laboratories and made available to all members of the research community (Figure 1). The RSSs will be used primarily to calibrate the internal standards and analytical methods used by individual laboratories to interrelate the doses used in different nonclinical and clinical studies. When reporting titers in the literature or to regulatory authorities, the relationship to an RSS could be included. The RSSs will be available in a form suitable for nonclinical and clinical data support, together with the profile and information regarding the development of each RSS.

Figure 1.

Sequence of events.

Full figure (20K)

The AAV2 Reference Standard Working Group (AAV2RSWG) was formed to produce and characterize an AAV2–green fluorescent protein (AAV2-GFP) viral vector RSS derived from the vector plasmid pTR-UF-11 (ref. 3). The AAV2RSWG is a volunteer organization comprising 12 industry members and 23 universities from nine countries, the US Food and Drug Administration/Center for Biologics Evaluation and Research (FDA/CBER), the National Institutes of Health (NIH), and the Williamsburg BioProcessing Foundation (WilBio, http://www.wilbio.com) (see Table 1). FDA and NIH representatives abstain during voting. The AAV2RSWG organized four committees to facilitate the effort: Manufacturing, Quality Control, and Donations, which are all coordinated by the Executive Committee. The working group has gathered several times over the past 5 years to plan and coordinate activities. WilBio has generously offered time and energy toward soliciting donations, posting information on the WilBio website, and coordinating meetings and conference calls as well as participating directly in the effort.

Table 1 - Reference standard participants.

Full table (134K)

Thus far, NIH National Gene Vector Laboratories funding has been used to establish this project ($80,611). The AAV2RSWG drafted requests for proposals for donations, and donations for raw materials and services (a value of $333,582) were delivered. Production and purification of the AAV2 RSS were carried out at the Vector Core of the University of Florida's Powell Gene Therapy Center, using helper virus–free transient transfection and chromatographic purification. Approximately 150 ml of AAV2-GFP = 5.69 10¹⁴ vector genomes (vg) (= 3.79 10¹² vg/ml) was made. The AAV2 RSS was diluted to the proper concentration (2 10¹¹ vg/ml), filtered, and divided into two 1.3 l bulk portions.⁴ After the vial label was finalized (December 2007), one of the 1.3 l portions of the bulk AAV2 RSS was divided by the American Type Culture Collection (ATCC) into 2,087 vials (March 2008) to produce ATCC VR-1616. Each vial of the AAV2 RSS contains 0.5 ml at a target dose of 1 10¹¹ vg. The vials, frozen in the repository at ATCC, await characterization and subsequent distribution. The other 1.3 l portion is being held and will be divided at a later date if demand warrants.

In parallel, members of the AAV2RSWG submitted their protocols to the Quality Control subcommittee for the various assays that will be used to characterize the AAV2 RSS. These protocols were reviewed (October 2007), and a lead protocol was chosen for each assay (November 2007). The assays include (i) confirmation of the serotype and capsid titer by A20 enzyme-linked immunosorbent assay (ELISA; Progen); (ii) evaluation of the purity, capsid subunit stoichiometry, and chemical integrity of the capsid by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE); (iii) vector genome titer by quantitative polymerase chain reaction (qPCR); (iv) determination of the empty:full capsid ratio (the ratio of A20 capsid titer to vector genome titer); (v) infectious titer by medium tissue culture infective dose (TCID₅₀) with qPCR readout and by transduction (GFP readout); and (vi) full sequencing of the vector. In addition, there will be a stability study that will comprise the infectivity and identity/purity assays performed at three separate time points (initial characterization, 6 months, and 1 year). The lead protocols were beta-tested by Martin Lock's group in the University of Pennsylvania's Gene Therapy Program (December 2007–March 2008), and the protocols were posted on the WilBio website (March 2008). In addition, the pTR-UF-11 plasmid was deposited (January 2008) and banked at ATCC and is available as ATCC no. MBA-331; the cell substrate for infectious titering (HeRC32 cells⁵) was deposited (January 2008) and banked at the ATCC and will be available as ATCC no. CRL-2972 following characterization. Vials of the adenovirus reference standard (ATCC VR-1516) have been made available for titering the AAV2 RSS (ATCC VR-1616), vials of which were distributed (May 2008) for characterization to the 23 laboratories that have volunteered to conduct single or multiple assays, followed by statistical analysis (August 2008), with availability expected in September 2008. The consensus titers and other characterization data from the multiple laboratories will be included on the ATCC product information sheet. The consensus data, by definition, will be recognized and accepted by the community, thus giving the RSS legitimacy.

Most of the 54 clinical trials to date have involved AAV2 vectors,¹ but vector systems based on other AAV serotypes are being rapidly developed. AAV8 is becoming widely utilized for gene transfer.⁶ Because the tropism of AAV8 is distinct from that of AAV2, different cells and organs can be targeted for transduction.

In Europe the effort has begun to generate an AAV8 RSS. Members of the AAV8RSWG have been assembled, including 3 from industry and 17 from academia, from 10 countries (see Table 1). Based on the model established for AAV2, a target of 2,000 vials filled with 0.5 ml at a concentration of 2 10¹¹ vg/ml of the AAV8 RSS will be produced by transient transfection, purified, and characterized. To harmonize the two reference standards, the AAV8 RSS will share the same vector genome derived from pTR-UF-11 that was used for the AAV2 RSS. Assay protocols, standards, controls, and cell substrates used for the AAV2 RSS will be shared for the characterization of the AAV8 RSS. Physical assays—vector genome titer (hybridization, PCR, or spectrophotometry) and SDS-PAGE—will be shared between the two reference standards. The characterization of the AAV8 will include the same parameters (confirmation of the serotype and capsid titer by ELISA using a conformational antibody (J. Kleinschmidt, personal communication)); evaluation of the purity, capsid subunit stoichiometry, and chemical integrity of the capsid by SDS-PAGE; vector genome titer by qPCR; empty:full capsid ratio; infectious titer by TCID₅₀ with qPCR readout and by transduction (GFP readout); full sequencing of the vector; and stability. Finally, there is considerable overlap between the two groups of volunteer laboratories participating (Table 1), and both efforts are being coordinated via a high level of communication.

The major difference with the AAV8 RSS is the need to obtain sufficient funding to support an accelerated effort. Generating the AAV2 RSS required the cumbersome solicitation of donations of time, materials, and laboratory space (totaling $414,500), a very lengthy process. Funding is being solicited to quickly establish the AAV8 RSS so that it can rapidly be made available in Europe and throughout the world.

Following the fine example set by the Adenoviral Reference Material Working Group for developing the adenovirus reference material (ATCC VR-1516),⁷ the members of the AAV community are developing two high-quality AAV RSSs. These will facilitate comparisons among nonclinical or clinical studies for classic single-stranded vectors and self-complementary vectors,⁸ aid in the manufacture of more consistent and higher-quality vectors, and ultimately help formulate regulatory policy.

In the coming months, the AAV2 RSS will be characterized, the data will be published, and the AAV2 RSS will be available through the ATCC. For more information on participation in the AAV2RSWG, please contact Richard Snyder; to join the AAV8RSWG, please contact Philippe Moullier.