RNAi as New Class of Nanomedicines - Harnessing Biomaterials in Nanomedicine: Preparation, Toxicity, and Applications - page 179

6.5 Off -Target Eff ects of siRNAs
Early RNAi-based gene-silencing experiments suggested that the
treatment of mammalian cells with synthetic siRNAs led to the highly
specific silencing of target gene expression without the induction of
non-specific interferon (IFN) responses [5, 16, 41, 67, 71]. In fact,
siRNAs could be identified that discriminated between transcripts
that varied by a single nucleotide polymorphism (SNP) attesting to
their high degree of specificity [5, 34, 41, 67, 129]. However, it has
become apparent that, under some conditions, the treatment with
siRNAs can have toxic eff ects. These toxicities can result from several
potential sources, including off -target silencing of genes that bear
partial complementarity to the siRNA, the engagement of innate
immune responses, and competition with endogenous RNAi eff ector
molecules, such as, miRNAs, for limited RNAi pathway components
[55, 73, 76, 126]. Recent microarray-based experiments have shown
that siRNA treatments can alter the expression levels of a large number
of transcripts. Informatic analysis has shown that these alterations
in transcript levels are sequence-dependent and largely the result
of the pairing of the 'seed region' (corresponding to nucleotides 2
to 8) of the siRNA to sites on off target mRNAs, particularly in the 3'
UTR, and therefore operating in much the same manner as miRNAs
[13, 75]. These “miRNA-like” off -target eff ects could be generated
from either strand of the siRNA duplex. The impact that these off -
target eff ects have on cellular functionality needs to be determined
experimentally and could potentially confound the interpretation
of experimental results. Since these eff ects are sequence-specific,
testing of multiple siRNAs against a given target mRNA can help to
delineate specific targeting from off -target eff ects. Alteration of the
thermodynamic properties of the siRNA ends to favor incorporation
of the guide strand could help to reduce off -target eff ects by reducing
the probability of incorporation of the passenger strand into the
RISC. Testing of multiple siRNAs targeting the gene of interest, each
with a unique seed sequence, is one approach that can be used to
ensure that the observed phenotype results from the knockdown
of the intended target and is not a result of the off -target silencing
[23]. In addition, rescue experiments can be performed by over-
expressing an siRNA-resistant, functional version of the target gene
to rescue the observed phenotype [36, 37].