Biology Reference
In-Depth Information
internalization or enhanced interaction with the intracellular RNAi machinery [ 158 ] .
Once inside the cytoplasm, siRNAs are more stable, some likely protected by RISC
incorporation [ 159 ], and silencing can persist for 30-90 days in slowly or nondivid-
ing cells [ 158, 160 ]. Therefore, chemical modification of siRNAs to enhance nucle-
ase resistance is primarily needed in applications using “naked” siRNA and less
when using shielding delivery agents such as nanoparticles.
1.4.5
Reducing siRNA Immunogenicity
Synthetic siRNAs were initially considered to be non-immunogenic in mammalian
cells, due to their structural mimicry of endogenous Dicer cleavage products [ 5 ] .
However, with increasing investigations, siRNAs are now reported to induce innate
immune responses through mechanisms dependent on cell type, delivery route,
siRNA structure, and sequence [ 76 ] . Speci fi c single-stranded sequence motifs such
as GUCCUUCAA [ 64 ] , UGUGU [ 129 ] , UGGC [ 161 ] , and GU [ 162 ] have been
reported to render siRNA immunogenic. These are recognized by Toll-like recep-
tors 7 and 8 (TLR7/8), transmembrane receptors found in the endosomes of immune
cell populations [ 163 ]. These responses can be potentiated with the use of transfec-
tion agents that facilitate endosomal delivery. Recently, uridine content has been
reported to correlate with TLR7/8 activation [ 164 ] , thereby severely complicating
siRNA design. Encouragingly, TLR7/8 activation may be largely avoided by chemi-
cally modifying or shielding immune-stimulatory sequences: Several modification
types, especially 2¢ -modi fi ed nucleotides (DNA, 2 ¢ -OMe, 2 ¢-F, LNA), can abrogate
siRNA immunogenicity [ 130, 147, 165, 166 ] , and modi fi cation of uridines only
with either 2¢-F or 2¢ -OMe [ 167 ] or DNA [ 168 ] may be suf fi cient. In particular,
2 ¢-OMe modification of the passenger strand has been proposed as a universal
approach to avoid TLR7 activation [ 169 ] .
siRNA duplex length also seems to affect siRNA immunogenicity; several stud-
ies suggest that 21mer siRNAs are not immunogenic in several cell lines even at
high concentrations, whereas 25mer (or longer) siRNAs trigger concentration-
dependent immunogenicity in HeLa S3, DU145, and MCF7 cells, but not in HeLa
and HEK293 cells. Consequently, it has been suggested to avoid non-modifi ed
27-29mer DsiRNA design for siRNA applications, at least in vivo [ 76 ] . siRNA
sequence length is sequence-independent monitored by the transmembrane TLR3
(among others) expressed on the surface and in endosomes in the dendritic sub-
population of the leukocytes but also in many primary cell types and popular cell
lines [ 76, 170 ]. Notably, TLR3 activation by the popular 21mer siRNA has been
reported upon intraocular injection in mice regardless of their sequence and 2¢ -
OMe modification, whereas a shorter 19mer siRNA design was safe [ 171 ] . Another
sensor of siRNA length is protein kinase R (PKR), present in all cells and stimu-
lated by cytoplasmic dsRNA longer than 30 bp [ 172 ] . 25-30mer siRNAs did not
activate PKR in HEK293 cells [ 71 ]; yet, canonical 21mer siRNAs have been
shown to bind or trigger modest PKR activation in murine microglial N9 cells
[ 173 ] , T98G cells [ 174- 176 ] , and HeLa cells [ 175 ]; therefore, the impact of PKR
Search WWH ::




Custom Search