Limitations of Encapsidation of Recombinant Self-Complementary Adeno-Associated Viral Genomes in Different Serotype Capsids and Their Quantitation

Abstract

We previously reported that self-complementary adeno-associated virus (scAAV) type 2 genomes of up to 3.3 kb can be successfully encapsidated into AAV2 serotype capsids. Here we report that such oversized AAV2 genomes fail to undergo packaging in other AAV serotype capsids, such as AAV1, AAV3, AAV6, and AAV8, as determined by Southern blot analyses of the vector genomes, although hybridization signals on quantitative DNA slot-blots could still be obtained. Recently, it has been reported that quantitative real-time PCR assays may result in substantial differences in determining titers of scAAV vectors depending on the distance between the primer sets and the terminal hairpin structure in the scAAV genomes. We also observed that the vector titers determined by the standard DNA slot-blot assays were highly dependent on the specific probe being used, with probes hybridizing to the ends of viral genomes being significantly overrepresented compared with the probes hybridizing close to the middle of the viral genomes. These differences among various probes were not observed using Southern blot assays. This overestimation of titer is a systemic error during scAAV genome quantification, regardless of viral genome sequences and capsid serotypes. Furthermore, different serotypes capsid and modification of capsid sequence may affect the ability of packaging intact, full-length AAV genomes. Although the discrepancy is modest with wild-type serotype capsid and short viral genomes, the measured titer could be as much as fivefold different with capsid mutant vectors and large genomes. Thus, based on our data, we suggest that Southern blot analyses should be performed routinely to more accurately determine the titers of recombinant AAV vectors. At the very least, the use of probes/primers hybridizing close to the mutant inverted terminal repeat in scAAV genomes is recommended to avoid possible overestimation of vector titers.

Absolute Determination of Single-Stranded and Self-Complementary Adeno-Associated Viral Vector Genome Titers by Droplet Digital PCR

“…However, findings of the underestimation of AAV self-complementary genome titers by qPCR (Fagone et al, 2011;Wang et al, 2012) led us to reexamine our AAV titering method and to explore the potential of digital PCR in AAV genome copy number determination. Our standard qPCR assay involves DNase I treatment of the sample followed by serial dilution, heat denaturation of the capsid to release the vector genome, and qPCR titration against a linearized plasmid standard curve (Fig.…”

“…1B; transgene and poly(A)] genome regions with either the ddPCR or the optimized qPCR assay, indicating that interference with primer-probe annealing was less pronounced at these genome positions. Furthermore, equivalent titers from assays targeting central and terminal genomic regions do not support the claim that packaged, terminal, defective interfering (DI) genome fragments artificially increase vector genome titers (Wang et al, 2012). It remains possible, however, that the assay targets were not entirely overlapped by the DI genomes.…”

“…Furthermore, different gel assay systems are required for ssAAV versus scAAV quantification, although data reported here suggest that this limitation could be overcome. Older methods of AAV genome quantification such as the DNA dot/slot blot and more recently, the Southern blot (Wang et al, 2012), can also overcome the problem of genome self-annealing but are particularly cumbersome and prone to operator-induced variability; they are therefore unlikely to be adopted as release assays in a commercial production environment. Direct quantification of AAV vector genomes by UV spectrophotometry (Sommer et al, 2003) is another assay, but is also limited to highly pure preparations.…”

Myosin7a Deficiency Results in Reduced Retinal Activity Which Is Improved by Gene Therapy

“…This 8.1 Kb expression cassette exceeds the canonical AAV cargo capacity and results in the generation of the so called “oversize” AAV vectors [27], [32], [33], [39], [40], [41], [42], [43], [44], [45]. Oversize AAV vectors have been reported to contain genomes of heterogeneous size [41], [42], [43], [44], [45], [46], [47] which may preclude their clinical application. Indeed, we have analyzed by alkaline Southern Blot the size of DNAse I-resistant viral genomes from oversize AAV2/5 vectors generated by using various pAAV2.1 cis -plasmids containing large transgenes (ranging from 7.7 to 8.9 Kb in size, see Materials and Method section).…”

“…However, AAV vectors have a cargo capacity considered to be limited to 4.7 Kb [41], [42], [43], [44], [45], [46], [47] and this is a major limitation when delivering large genes like MYO7A . This can be overcome by either generating oversize AAV vectors by using a plasmid containing a large gene expression cassette [27], [32], [33], [39], [40], [41], [42], [43], [44], [45] or by splitting a large gene into two halves each contained in a separate, regular size AAV vector (dual-AAV vectors [59]).…”