Evidence for the failure of adeno-associated virus serotype 5 to package a viral genome > or = 8.2 kb

Abstract

Limited packaging capacity hinders adeno-associated virus (AAV) gene therapy. A recent study seems to have provided a solution to this problem. Allocca et al. reported that AAV-5 could package an 8.9 kb vector genome. Here we tested whether this approach can be used to deliver a large genome for Duchenne muscular dystrophy (DMD) gene therapy. We first evaluated AAV-5 packaging of an 8.2 kb genome. This vector carries two independent reporter gene cassettes, one for alkaline phosphatase (AP) and another for LacZ. Viral yield was log-fold lower than that of a regular AAV-5. Nevertheless, both AP and LacZ genes were detected in purified virus. Injection to dystrophic muscle resulted in both AP and LacZ expression. On electron microscopy, virion structure appeared normal. Surprisingly, we did not find the full-length single-stranded viral genome by alkaline gel electrophoresis. Neither did we see the full-length double-stranded replication forms in adenovirus coinfected cells. We suspect that AP and LacZ expression may have come from partially packaged 5' or 3'-half of the genome. Additional studies revealed failure of AAV-5 to package and express an 8.7 kb minidystrophin gene cassette. In summary, our results do not support the extraordinary packaging capacity of AAV-5.

Figure 1

AAV-5 packaging of an 8.2 kb viral genome. (a) Schematic outline of the putative full-length viral genome. The 5′-half genome contains an AP expression cassette and the 3′-half genome contains a LacZ expression cassette. RSV, Rous sarcoma virus promoter; pA, SV40 polyadenylation signal; AP, alkaline phosphatase gene; LacZ, β-galactosidase gene. The locations of the AP and LacZ probes are marked. (b), A representative slot blot of the purified AV.AP.LacZ virus. vg, viral genome.

Figure 2

Characterization of AV.AP.LacZ infection in dystrophic muscle. Representative photomicrographs of HE, AP, and LacZ stained muscle sections. Transgene expression was detected at 1 month after AV.AP.LacZ infection (N = 4 muscles). Bottom panels are high magnificent photomicrographs of the boxed areas in the respective low power images. Asterisk, a myofiber with high level AP expression also showed high LacZ expression; Cross, a myofiber with relatively weak AP expression also showed minimal LacZ expression; Arrow, a myofiber with robust AP but poor LacZ expression; Arrowhead, a myofiber with relatively high level LacZ but weak AP expression.

Figure 3

Characterization of AV.AP.LacZ viral particle and the vector genome. (a) Representative electron microscopic images of AV.AP.LacZ and AV.EGFP viral particles. AV.EGFP carries a 4.7 kb vector genome. (b) A representative alkaline gel Southern analysis of the AV.AP.LacZ viral genome. (c) A representative Southern blot examination of the replication forms of AV.AP.LacZ virus. The probes used are marked for each blot. Results in each panel were obtained from at least three independent experiments.

Figure 4

AV.ΔH2-R15 fails to express the minidystrophin gene in mdx muscle. (a) Schematic outline of the putative full-length AV.ΔH2-R15 genome. The locations of the Dys-3 and Dys-2 antibodies are marked. (b) Representative photomicrographs of serial muscle sections stained with a human dystrophin hinge-1 specific Dys-3 antibody or a dystrophin C-terminal domain specific antibody (Dys-2). Arrow, rare occurring revertant fibers detected by the Dys-3 antibody. Arrowhead, infiltration of immunoglobulin in necrotic myofiber.

Figure 5

AAV-5 packaging of normal and large sized viral genomes. (a) Schematic presentation of AV.AP.LacZ (left panel) and AV.ΔH2-R15 (right panel) packaging. Middle panel depicts AAV-5 packaging of a wild-type (wt) sized genome. AV.AP.LacZ and AV.ΔH2-R15 are consisted of particles carrying either the 5′-half or the 3′-half of the vector genome. In AV.AP.LacZ, independent packaging of each expression cassette by individual viral particle resulted in AP and LacZ coexpression in the same muscle cell. In AV.ΔH2-R15, failure to express minidystrophin is most likely due to an incomplete packaging of the full-length cassette in a single virion. Also included are a representative histochemical and dystrophin immunostaining images from AV.AP.LacZ or AV.ΔH2-R15 infected mdx muscle. The immunostaining image from normal muscle reveals sarcolemmal dystrophin expression. N, the N-terminal end of minidystrophin; C, the C-terminal end of minidystrophin. (b) Transgene expression from AAV-5 carrying different size viral genomes. For virus with a normal sized genome, both the 5′ and the 3′-end inverted terminal repeats (ITRs) are encapsidated in mature viral particle. For these with an oversized genome, we hypothesize that there is only one ITR in each mature virion. For a large but highly recombinogenic target gene, we speculate that transgene expression may have derived from homologous recombination between the overlapping regions of partially packaged viral genomes.