AAV Gene Therapy Strategies for Lysosomal Storage Disorders with Central Nervous System Involvement - Gene Delivery and Therapy for Neurological Disorders - page 269

Fig. 2 Genomic organization of recombinant adeno-associated viral vectors
to date, there are no reports linking it to any disease [ 54 ].
Recombinant AAV (rAAV) vectors are devoid of viral coding
sequences and carry an ITR-fl anked transcription cassette consist-
ing of a promoter driving expression of a gene of interest and a
polyadenylation signal (polyA) (Fig. 2 ). The ITR elements are
necessary for vector genome replication and packaging during
production and postinfection processing to generate stable tran-
scriptionally active genomes in target cells.
Recombinant AAV has become the platform of choice for
in vivo gene delivery to the CNS for a number of reasons including
its ability to transduce dividing and nondividing cells at high effi -
ciency, mediate long-term gene expression, and maintain an excel-
lent safety profi le both in animal models and human clinical trials.
However, the immunology of AAV-encoded transgenes after gene
transfer to the CNS is an evolving fi eld, and it appears that new
more powerful capsids, such as AAV9, are also more likely to trigger
immune responses than less robust AAV serotypes, such as AAV2,
when encoding either nonmammalian transgenes (GFP) or species
mismatched therapeutic genes [ 55 ]. Most experiments in mouse
models of LSDs have been conducted with AAV vectors encoding
human enzymes. In some animal models, the therapeutic effi cacy
appears to be compromised by the immune system [ 56 , 57 ], and it
is unclear whether this is due to an immune response to a species
mismatched enzyme or because of the knockout nature of most
mouse models in which no protein is detectable. This aspect should
be considered carefully in the planning of preclinical effi cacy and
safety experiments in large animal models as the use of AAV vectors
encoding human enzymes can confound the interpretation of out-
comes [ 58 ]. Although the degree of homology between human
and other mammalian enzymes (especially monkey) is very high,
there may be suffi cient amino acid differences to generate new epi-
topes that may trigger an immune response. The approach that is
often employed to test AAV vectors encoding human enzymes in
large animals is to use immunosuppression [ 59 , 60 ]. For obvious
reasons, this can blunt or mask responses that would otherwise
become apparent and possibly limiting to the implementation of a
particular vector design and/or delivery route. Finally, the amino
acid differences between enzymes from different species could alter
protein-protein interactions in subtle ways and infl uence both ther-
apeutic outcomes and toxicological profi les. For these reasons, we