AAV-mediated gene delivery remains a potential treatment option for some patients with muscular dystrophy. Challenges, however, for DYSF gene replacement relate to AAV packaging limits for genomes significantly larger than wild-type length (4.8 kb). Fortunately, the sentinel report of Alloca et al.
brought enhanced packaging by AAV5 to light and gave new hope for expanding the capacity of AAV gene transfer for monogenic disorders of large genes with the potential to be carried to the bedside 
. In their studies, efficient packaging, transduction, and expression of an 8.9 kb cassette of a large inherited gene causing blindness implied the ability to overcome the hurdle of a 5 kb size limitation for AAV for replacement therapy. Our current study along with three others 
have demonstrated that homologous recombination of partially packaged genomes is the mechanism responsible for the generation of full-length transcripts rather than oversized packaging of the whole genome. Regardless, AAV5 does exhibit enhanced plasticity regarding packaging constraints which is likely contributing to its ability to mediate production of full-length transcripts in muscle following homologous recombination 
. With regard to neuromuscular diseases, the findings provided a new perspective for conditions caused by mutations of large genes. DMD is the most common severe childhood muscular dystrophy and would seem to benefit from expression of the larger transcripts than mini- and micro-dystrophins that only partially restore physiologic function in the mdx mouse 
. Less common disorders, such as titin deficiency causing LGMD2J, and variants of congenital muscular dystrophy such as phenotypes caused by LARGE gene mutations would also benefit from expression of the full length protein 
. Because of the success of Alloca et al.
, it was our intent to take advantage of the transduction capabilities of AAV5 for skeletal muscle 
. We targeted the dysferlin gene because of its relative frequency amongst the LGMDs 
. Our findings are supportive of the unique properties of AAV5 to express a 6.5 kb cDNA producing full length protein but the mechanism for expression of the dysferlin gene pointed in a different direction than reported for the series of retinal genes protecting against blindness 
. Our studies confirmed that AAV5 followed traditional packaging limits for DYSF because analysis of packaged virions was consistent with a packaging limitation of ~5 kb. Thus, AAV5 virions transducing muscle fibers contained partial dysferlin vector genomes that mapped to both the 5′ and 3′ ends of the expression cassette and produced an intact expression cassette likely through homologous recombination upon reaching the nucleus 
. This enabled translation of the full length 237 kDa protein demonstrated by western blots following gene transfer. The findings were equally robust following intramuscular and intravascular delivery, and the validity of the newly expressed protein was confirmed by its full restoration of physiologic function in the diaphragm and membrane repair capabilities in skeletal muscle.
Of particular importance, the potential for DYSF gene replacement has also been demonstrated by Lostal et al.
with their comparable findings demonstrating expression of full-length dysferlin 
. These researchers used a dual vector strategy permitting the dysferlin cDNA to be split at the exon 28/29 junction and cloned it into two independent AAV vectors carrying the appropriate splicing sequences. Their work tested the efficiency of dual packaging into AAV2/1 by intramuscular injection of both vectors into a dysferlin-deficient mouse and a novel strategy using systemic delivery of dual vector administration of both rAAV2/1 and rAAV2/9. Overall, both IM and systemic vascular delivery via the tail vein of dysferlin-deficient mice showed improvement of muscle pathology. Their study also noted an improvement in membrane repair in the FDB that did not reach WT control levels. By systemic delivery they found low levels of transduction (1–4%) which is a current limitation of the dual vector approach. In a clinical setting, higher levels of gene expression may be required to achieve a clinically meaningful outcome. Lostal et al
suggested that 30% levels of dysferlin may be needed (26).
An additional report by Krahn et al.
described a human minidysferlin protein that was identified in a patient with late-onset moderate dysferlinopathy 
. The 73 kDa protein lacks the consensus dysferlin N-terminus but maintains the wild-type C-terminus including the last two C2 domains and transmembrane domain. The authors packaged this minidysferlin cDNA in a cassette containing a C5.C12 promoter and delivered it to A/J dysferlin deficient mice using AAV2/1 and AAV2/9. Expression of the protein was confirmed in the TA and found to localize primarily to the cytoplasm of muscle fibers and to a lesser extent in the sarcolemma. An examination of function revealed that the minidysferlin led to partial improvement in the membrane repair deficits accompanying the dysferlin-null FDB muscle fibers 
Our studies demonstrate clinical applicability for AAV5.DYSF gene transfer for dysferlin deficiency. There are several advantages to using AAV5 to deliver the full-length dysferlin cDNA. One particular advantage is the ability to deliver the entire cassette with one vector, significantly reducing viral load compared to the dual vector strategy. This is especially important with regards to translation in terms of feasibility of vector production and limiting capsid exposure for patients. A second advantage relates to full-length protein versus a miniature version. Although mini-genes are desirable for practicality of standard AAV delivery, there is a compromise in protein function 
and the potential for novel epitopes that may be immunogenic when delivered to patients 
. One potential concern with AAV5.dysferlin delivery is the presence of non-recombined vector genome fragments; however we showed only one mRNA and protein species was present in transduced muscle. As with any other naturally occurring truncated transcript that could be produced; elimination would occur by nonsense mediated decay. An informal discussion with the FDA (Rodino-Klapac and Mendell personal communication) defined a potential path for a dysferlin clinical gene therapy trial assuming no problems were encountered in the toxicology/biodistribution studies done with the same rigor as other approved AAV vectors 
In conclusion, we have shown that AAV5.dysferlin delivery is a very promising therapeutic approach that could restore functional deficits in dysferlinopathy patients. Specific muscle groups could be targeted by intramuscular delivery for dysferlin phenotypes that include Miyoshi myopathy and distal anterior compartment myopathy. In addition, based on our experience using fluoroscopy guided vascular delivery studies in the non-human primate we can thread the intravascular catheter to the take off point of specific vessels 9. This opens up clinical pathways for gene delivery to particular muscle groups relevant to either LGMD2B or Miyoshi myopathy including quadriceps and hamstring muscles, and the anterior or the posterior compartments of the lower limb.