RNAi Therapeutic Delivery by Exosomes - RNA Interference from Biology to Therapeutics - page 196

methods, maintained in culture or frozen for much greater lengths of time than
primary DCs, and differentiated into a chosen lineage, e.g., neurons and DCs
[ 82- 84 ]. As well as eliminating immunogenicity, this approach offers the option of
using exosomes from stem cell-derived neurons or other brain cells, which are likely
to display intrinsic neurotropic behavior and enhanced brain specificity. In addition,
the applicability of exosome-delivered therapy to a large number of disease condi-
tions would be greatly enhanced by broadening their cargo repertoire and improving
the tissue-targeting strategies. Indeed, it would be important to explore other car-
goes such as miRNA, miRNA antagonists, and shRNA expression plasmids. The
delivery of miRNA mimics and antagonists by exosomes would be particularly
beneficial where multiple pathways are targeted, as is the case in Alzheimer's
(miR29) [ 85 ] and in cancer (e.g., tumor suppressor miR-7 and miR-128 replace-
ment therapy in glioblastoma [ 86, 87 ]). These miRNAs would either be loaded
directly into exosomes or enriched in them by expression in the exosome-producing
parent cell providing a continuous source of loaded exosomes. The identification of
novel targeting moieties, other than RVG, specific for the brain or other tissues of
interest will further broaden the therapeutic applications of exosomes. Attractive
candidates include monoclonal antibodies against receptors that are naturally
expressed on the BBB or adhesion molecules expressed on endothelial cells in the
lining of blood vessels. It would, therefore, be possible to achieve a degree of tissue
specificity not previously achievable with viruses and liposomes, further enhancing
the bioavailability of exosome cargo, reducing nonspecific homing to other tissues
and to sites of clearance. These targeting moiety or moieties could either be
expressed in the parent cells or inserted directly onto the exosomal membrane,
allowing to control for levels of surface expression and the relative ratios of target-
ing moieties.
In addition to harnessing exosomes for RNAi delivery, lipidomic, proteomic, and
transcriptomic studies promise to define more precisely the components that make
exosomes competent in drug delivery. Defining these components will aid the
design and development of (semi-)artificial drug delivery vehicles. Indeed, an alter-
native approach to obtaining membrane vesicles is to generate them synthetically
by recapitulating the components that are essential for their function as a carrier
system. Such synthetic exosome mimics would be a homogenous and reproducible
type of drug delivery vehicles, devoid of other cargoes that are naturally found in
exosomes and which otherwise mediate off-target effects. To mimic the exosomes'
natural targeting properties and their ability to deliver RNAi to the site of tissue
silencing, it is likely that artificial particles, e.g., liposomes, would have to comprise
the specific lipids and proteins involved in exosomal cell trafficking. An attempt to
produce such “exosome-like” vesicles has already been described by Martinez-
Lostao et al. [ 88 ]. The effectiveness of APO2L/TRAIL (proteins that are decreased
in the synovial fluid from patients with rheumatoid arthritis) conjugated with
artificial lipid vesicles was evaluated in a rabbit model of antigen-induced arthritis
(AIA) using artificial exosome-like liposomes which included phosphatidylcholine
(PC) and sphingomyelin and were conjugated to APO2L/TRAIL. The authors