Mucosal Delivery of RNAi Therapeutics - RNA Interference from Biology to Therapeutics - page 107

Whilst PEGylated PEG-PEI polymers appear less cytotoxic than non-modified
PEI [ 104 ], they and the fatty acid modified PEI may have a higher proinflammatory
potential that is of concern in a clinical setting. Although PEI has been used exten-
sively for several years in animal studies, safety concerns still restrict its use in the
clinic.

5.3.3.2 Lipid-Based Systems

Lipid vectors have been widely used for in vitro and in vivo delivery of siRNA
[ 105 ], most based on cationic lipids that form lipoplexes with siRNA. Mirus TKO,
a cationic lipid/polymer formulation, has been used by Bitko et al. to deliver siRNA
against RSV [ 63 ] . 70 mg siRNA delivered intranasally with Mirus TKO was able to
reduce viral titres in mice without inducing an interferon response, an effect shown
to be a 20-30% improvement on naked siRNA. In a mouse influenza model, the
animals received hydrodynamic injections (3.78 nmol) of naked siRNA followed
16-24 h later by intranasal delivered oligofectamine/siRNA (1.51 nmol) complexes
against the viral nucleoprotein and acidic polymerase to reduce viral titres in the
lung (63-fold compared to EGFP siRNA) [ 71 ]. Whilst interferon levels were inves-
tigated and found not to be upregulated, concerns remain for the use of the EGFP
sequence as a negative control due to its non-stimulatory uniqueness [ 101 ] . A third
commercial lipid-based transfection reagent, DharmaFECT, has been used as a pul-
monary delivery vector in a bleomycin-induced lung fibrosis mouse model [ 72 ] .
siRNA against SPARC, a matricellular protein overexpressed in fibrotic diseases,
markedly reduced collagen content in the lung (58%) compared to the bleomycin-
only group after intratracheal dosing (3 × 3 m g siRNA).
The cationic lipid from Genzyme, GL67, has been used in k18-lacZ mice which
express b-galactosidase in airway epithelial cells [ 106 ]. A 33% reduction of mRNA
level (but no change in protein levels) was observed after intranasal administration
of lacZ siRNA (40 mg siRNA). Histological analysis showed that the GL67/siRNA
lipoplexes were mainly associated with pulmonary macrophages which could
explain the lack of change in protein levels.
Direct conjugation of siRNA to cholesterol has been explored by Moschos et al.
[ 73 ]. A single intratracheal administration of siRNA-cholesterol conjugates
(10 nmol) facilitated a 45% knockdown of p38 MAP kinase mRNA in mouse lungs
after 12 h compared to vehicle-only controls. The effect appeared to be transient as
the detected mRNA levels were almost back to normal after 24 h which was attrib-
uted to poor stability of the siRNA. It was suggested that chemical modification of
the backbone might increase the silencing effect.
The respiratory vasculature in mice can be targeted by systemic delivery of cat-
ionic lipoplexes (AtuPLEX) [ 74 ]. A ~50% reduction of the endothelial cell-specific
protein VE-cadherin was achieved in the lungs after intravenous injection of 50 m g
of siRNA on 4 consecutive days compared to a luciferase-specific siRNA. The lipo-
plexes were also administered intratracheally, but only a 21% reduction of epithelial