Mucosal Delivery of RNAi Therapeutics - RNA Interference from Biology to Therapeutics - page 106

administration [ 4 ] suggests circulatory properties. Furthermore, modification of
chitosan with an imidazole group and PEG has been used for intravenous delivery
of siRNA in mice resulting in a 49% reduction of mRNA levels of GAPDH in the
lungs, suggesting that chitosan/copolymers might be useful as an intravenous deliv-
ery vector [ 67 ] .
PEI-Based Systems
Since the introduction of polyethylenimine (PEI) as a gene transfer reagent in 1995
[ 99 ], this cationic polymer has been studied extensively for both DNA and siRNA
delivery [ 100 ]. Effective polyplex formation, protection from nucleases and endo-
somolytic properties attributed to its high charge amino density have promoted its
use. PEI has been used for systemic delivery of siRNA to a number of tissues includ-
ing the lungs in mice [ 68, 69 ]. Ge et al. and Thomas et al. used PEI/siRNA poly-
plexes (N:P 5) against the influenza nucleocapsid protein (120 mg single dose)
administered by retroorbital injection. The study by Thomas et al. expanded on the
previous study by Ge et al., by evaluating the ability of various high molecular
weight PEI polymers to enhance the delivery of siRNA and in both studies, a
significant reduction of viral titres was observed (10- to 1,000-fold reduction).
The mechanism of antiviral effects from the studies by Thomas and Ge, however,
has been brought to question in a landmark paper from Robbins et al. [ 101 ] . In this
work, several published siRNA sequences, including the nucleocapsid sequence
used by Ge et al. and Thomas et al., were tested for an ability to stimulate the innate
immune system ascribed to intracellular Toll-like receptor activation. Remarkably,
it was shown that the control GFP sequence used in several in vivo studies [ 60, 68,
69, 71 ] did not elicit an immune response, whereas the nucleocapsid sequence
(among others) stimulated the production of interferon a suggested to be largely
responsible for the observed antiviral effect.
The ability of systemic PEI-based siRNA systems to reach the lungs [ 102, 103 ]
could result from serum-induced aggregation and its consequent entrapment within
the lung capillary beds. This, however, could result in lung embolisms and restricts
the likelihood for clinical translation.
PEI, unlike chitosan, is not generally thought to be a mucoadhesive polymer, but
it is reasonable to speculate that some amino-mediated interaction with mucins can
occur if delivered locally. Hitherto , PEI interactions and effects on mucus have not
been studied in detail. Two recent studies have demonstrated pulmonary EGFP
silencing in transgenic mice. Using intratracheal administration of PEG-PEI/siRNA
polyplexes, Merkel et al. showed a 42% reduction of EGFP expression compared to
luciferase siRNA control (single 50 mg siRNA dose) [ 57 ] . Moderate in fl ammation
was seen by analysis of cytokine levels, but no histological abnormality was observed.
Beyerle et al. used a fatty acid modified PEG-PEI/siRNA polyplex to achieve a 69%
reduction of EGFP expression compared to untreated controls (35 mg single dose)
[ 70 ]. As in the previous study by Merkel et al., PEGylation increased inflammation,
whilst at the same time also decreased cytotoxicity. These findings seem to contra-
dict the usual perception that PEG limits interaction with the immune system.